The temporal-spatial pattern of linear cultural heritage in the context of the tourism industry is closely linked to heritage management. Using the 1800 km long Beijing-Hangzhou Grand Canal as an example, this study compared the dynamic evolution of tourism businesses in Beijing, Liaocheng, and Yangzhou at three time points (2010, 2015, and 2019) via nearest neighbor analysis, kernel density estimation, and the standard deviational ellipse. Next, a Geo-detector was used to examine the influencing factors. The results reveal significant growth regardless of the quantity or agglomeration degree from 2010 to 2019, and the direction of industrial expansion is consistent with the flow direction of the canal. Moreover, the explanatory powers of factors related to socioeconomic development and canal resources are obviously stronger than those of the natural environment. The findings of this study offer theoretical constructs and policy recommendations for the sustainable development of the Beijing-Hangzhou Grand Canal and other linear cultural heritage sites.
1 Introduction
Linear cultural heritage (LCH) is a new type of world heritage. It reflects inter-regional exchanges and human activities in a linear or strip geographical space during a specific period, such as military fortifications, water conservancy and transportation projects, and trade and religious routes (Shan, 2006; Yu et al., 2009). Since the end of the 20th century, LCH has gradually become a concern for governments, academics, and the public worldwide. Unlike dot cultural heritage, such as individual cultural relic buildings, or planar cultural heritage, such as towns or village landscapes, LCH usually involves multiple administrative districts, cultural regions, and economic zones (Zhang et al., 2020), so it has the characteristics of a complex structure, high interactivity, a wide spatial span, and strong cultural continuity (Li and Zou, 2021).
Implemented as an effective solution for displaying local culture (Wu and Wang, 2018), tourism offers feasible paths for LCH protection and activation. However, due to unbalanced resource distribution and complex spatial composition, LCH faces a significant dilemma for the different destinations along heritage sites, with the number of visitors being either too high or too low. By providing visitors with necessary services such as dining, lodging, transportation, sightseeing, and shopping (Aratuo and Etienne, 2018), travel-related businesses directly present an LCH's current situation and potential as a destination. Understanding the temporal-spatial distribution of the tourism industry can clarify the development models and supporting conditions of the LCH, which is significant for different areas along the LCH according to the actual situation, and avoids inappropriate development and management.
The world's largest and longest manmade canal, the Beijing-Hangzhou Grand Canal, stretches nearly 1800 km across 22 cities that are at or above the prefecture level in eastern China. This canal is an outstanding ancient water conservancy and shipping engineering technology (Peng, 2014). However, similar to other LCHs, the Beijing-Hangzhou Grand Canal is also experiencing the problem of unbalanced tourism development. Exploring the temporal-spatial distribution pattern of the tourism industry along the canal can promote the positive interaction between heritage and leisure activity, and provide theoretical guidance on heritage management and tourism marketing for LCH worldwide.
This study's aforementioned rationale is applied in the pursuit of two objectives: 1) Compare the temporal-spatial distribution of the tourism industry along the Beijing-Hangzhou Grand Canal and 2) Analyze the factors influencing the tourism industry distribution. Based on these two objectives, this study contributes to the theoretical advancement of LCH management and leisure marketing from an internal comparative perspective and can guide feasible policy formulation in a sustainable manner.
2 Literature review
2.1 Tourism utilization of linear cultural heritage
LCH represents a deep understanding of the continuous protection of cultural heritage after extending and integrating similar concepts, such as “cultural route”, “heritage corridor” and “heritage route” (Lu, 2014). Compared with these concepts, LCH has a broader definition standard, which can be explained as a subcategory of cultural heritage formed by spatial morphological constraints (Harvey, 2015). This study defines LCH as a cultural heritage group that relies primarily on linear spaces such as rivers, canyons, traffic lines, and commercial roads, with outstanding thematic and functional features. Studies on LCH have mainly discussed their functional evolution (Oviedo et al., 2014), value assessment (Božić and Tomić, 2016), spatial structure, and protection and utilization, and have mostly focused on specific administrative regions, or heritage areas with definite boundaries (Hoşgör and Yigiter, 2011). However, cross-regional comparative studies within the LCH should be further explored (Caton and Santos, 2007).
Tourism and leisure utilization are effective approaches for cultural heritage activation (Wu and Wang, 2018). Considering the high degree of integration in the tourism and leisure industry (Zhang et al., 2021), this study focuses on the extensive role of host and guest sharing; thus, it does not distinguish between the two tourism products (Gravari-Barbas, 2018), brand building (Hou and Zhang, 2019), or effects (Campolo et al., 2016), all of which promote the diversity and richness of LCH tourism and leisure research. However, because of the outstanding universal values of LCH, most studies have overemphasized the heritage and paid insufficient attention to the external visiting environment that is related to economic and social development (Tuxill et al., 2008). As a result, although the cultural connotation of LCH has been demonstrated, the driving effect of the tourism industry has not been fully explored. Tourism and leisure are derivative functions of LCH and a complex system containing many elements (Telfer, 2001). Exploring the utilization pattern and evolutionary law of the tourism industry along the LCH is conducive to evaluating the development potential and promoting comprehensive benefits, especially from the perspective of the coupling of the heritage and tourism industries.
2.2 Spatial-temporal distribution of the tourism industry along linear cultural heritage
The temporal-spatial distribution illustrates the evolving process and characteristics of tourism development, which helps to clarify the layout rationale of the supply market and provides guidance for the scientific management of LCH (Zhang et al., 2020). Because of the large geographical span and complex regional situation, the temporal-spatial distribution of the tourism industry along an LCH is mainly discussed in two aspects. On the one hand, the integrity of LCH determines that most studies explore its evolutionary laws from the perspective of the heritage as a whole (Murray and Graham, 1997). On the other hand, some studies have focused on partial sections or regions of an LCH to support dynamic updates of localized marketing and management for the large-scale heritage (Ren, 2017).
Based on empirical analyses, scholars have proposed the “node-channel-network” and “node-link-proposition-hierarchy” spatial modes, which emphasize the key position of nodes in LCH tourism (Wang et al., 2012). Temporal changes in distribution can determine the reasons for the evolution of the tourism industry by distinguishing different developmental stages and facilitating the prediction of future trends (Oviedo et al., 2014). Studies have generally regarded the spatial distribution of LCH tourism and leisure utilization as an evolving process. Although an increasing number of studies are available on the temporal-spatial distribution of the tourism industry of LCH, most have focused on the whole heritage or only partial sections.
2.3 Factors influencing the tourism industry distribution along linear cultural heritage
Due to the various social, cultural, and political backgrounds, the geographical distribution of the LCH tourism industry tends to display diverse situations (Ren, 2017). Heritage resources, tourist attractions, traffic conditions, cultural connotations, ecological environments, service facilities, tourist markets, policies, and regulations are common factors influencing the spatial layout of the LCH tourism industry (Hoşgör and Yigiter, 2011; Draper et al., 2016; Gravari-Barbas, 2018; Zhou, 2021). The large number of factors illustrates that the tourism industry is a comprehensive system and closely bound to the heritage itself and the surrounding environment (Zhang et al., 2020). A summary of the existing studies shows that three categories of factors have been mentioned frequently. First, tourism businesses usually present spatial agglomerations along the LCH, and spread out and decline to both sides (Zhang et al., 2021). As the core tourist attractions, heritage resources directly affect the spatial location of tourism businesses (Zhang et al., 2023a). Second, LCH tourism is highly dependent on the natural environment (Zhang et al., 2023b). The prosperity of tourism in different sections along an LCH is significantly different under the influence of natural factors such as air quality and greening rate (Zhang et al., 2023a). Finally, the tourism industry is deeply rooted in the local socioeconomic environment, and economically developed sections along the LCH will stimulate tourism market vitality and build well-known tourism brands (Zhou, 2021).
The temporal-spatial heterogeneity along the LCH tourism industry is the result of many factors, so revealing the influencing factors can further clarify the industrial layout principle and provide guidance on tourism marketing and destination management for practical application.
2.4 Tourism development of the Beijing-Hangzhou Grand Canal
Beijing-Hangzhou Grand Canal is a vast inland waterway system, stretching from Beijing in the North China Plain to Hangzhou in the Middle and Lower Plain of Yangtze River. Constructed in sections from the 5th century BC onward, it was regarded as the world's largest and most extensive hydraulic engineering project prior to the Industrial Revolution. Relying on diverse cultures and resources such as water conservancy, commercial trade, the royal system and local literature, there are numerous and distinctive tourism resources along the canal (Li and Zou, 2021).
Most studies on the tourism industry of the Beijing-Hangzhou Grand Canal are combined with resource evaluation and development potential, involving different spatial scales such as cities, regions or the whole canal (Wang, 2018). However, like other LCHs, this nearly 1800 km canal also has the problem of unbalanced development, and the tourism industry is no exception (Peng, 2014). Hence, comparing and distinguishing the different development conditions and influencing factors of the tourism industry is necessary for this canal.
3 Methodology
3.1 Study sites
Considering the heritage function and tourism development, this study selected Beijing, Liaocheng and Yangzhou for the comparative analysis (Fig. 1). Specifically, although the original functions of canals, such as shipping, were barely operational in Beijing, many recreational spaces along the canal have been built to stimulate tourism activities. Canal recession also happened in Liaocheng. With the watercourse drying up or even disappearing, the heritage resources and tourism attractions are insufficient, and the tourism market development has lost vitality. Unlike the canal in the first two cities, the canal in Yangzhou has always played a significant role in shipping and water conservancy. As a canal city, Yangzhou is also one of the most prominent tourist destinations along the Beijing-Hangzhou Grand Canal. In addition, these three cities are located in the northern, middle, and southern portions of the canal, respectively, which benefits the comparison of differences along the entire canal. Therefore, this study considers different combinations of heritage and tourism to explore the issues related to the tourism industry using the most representative sections along the canal.
3.1.1 Beijing section
Although the Beijing section of the canal has not fully realized its intended function, as an urban ecological recreational space, it is receiving increasing attention. The government has strengthened relevant regulations, with participation from the community. The Beijing Section of the canal has been a model for improving water quality and preventing pollution. Since its transformation, the Canal mainly serves to improve the urban ecological environment and it plays a notable role in the urban leisure and recreation (Table 1).
3.1.2 Liaocheng section
The Liaocheng section of the canal is in poor condition and entirely silted up, so it only offers surrounding water conservancy facilities, ancient buildings, inscriptions, and so forth. The canal flowing through the downtown area of Liaocheng became a channelized river after its renovation and is designed as a waterway for urban canal cruises (Table 1). However, there are many abandoned canals in the rural areas. Although the government has conducted canal restoration, villagers remain unaware of the value of the canals. Notably, daily sewage and contamination in rural areas can lead to serious water pollution.
3.1.3 Yangzhou section
The Yangzhou section is a well-protected section of the Beijing-Hangzhou Grand Canal, especially the ancient canal within the urban area. As the main landscape belt and recreational space of Yangzhou, this section contributes to urban greening, leisure activities for citizens and sightseeing for tourists (Table 1). Based on the ancient canal, the main tourist business district of Yangzhou was formed by actively exploring the canal culture related to, for example, canal cruises, bank landscapes, and cultural relics.
3.2 Data collection
3.2.1 Tourism industry data
This study mainly referred to the National Statistical Classification of Tourism and Related Industries (2018) to divide the tourism industry into five categories: tourism landscape (e.g., tourist attractions or scenic spots related to sightseeing, recreation, vacation, and folk culture), tourism shopping (e.g., shopping centers, large shopping malls, commercial blocks, and tourist souvenir shops), accommodation facilities (e.g., star-rated hotels, hostels, resorts, and homestays), leisure catering (e.g., restaurants, snack bars, bars, teahouses, and coffee shops), and entertainment (e.g., theaters, swimming pool complexes, gymnasiums, and cultural exhibition halls). Three time points (2010, 2015, and 2019) were selected for comparison to show the temporal-spatial evolution of the tourism industries in different cities. These time points were selected mainly because The Grand Canal was listed as a World Cultural Heritage in 2014. Considering that the brand effect of heritage has a huge impact on tourism, with 2015 as the first year after becoming a worldwide heritage, it basically reflects the status of applying for World Cultural Heritage. The years of 2010 and 2019 respectively reflect the development of the similar times before and after that notable change.
First, the statistics of the tourism industry in Beijing, namely, name, location, and opening time, were collected from the Beijing Statistical Yearbook, Development Plan of Tourism and Exhibition Industry during the 13th Five-Year Period of Beijing, Development Plan of Total Number and Layout of Entertainment Venues in Beijing during in 2013–2015, Statistical Communiques and Annual Work Reports of the National Economic and Social Development, and the official websites of the Beijing Municipal Bureau of Statistics ( http://tjj.beijing.gov.cn/tjsj/), the Beijing Municipal Bureau of Culture and Tourism ( http://whlyj.beijing.gov.cn/), and BIGEMAP software. Next, the databases of Liaocheng and Yangzhou were formed separately in the same manner. The tourism industry data of Liaocheng were collected from the Liaocheng Statistical Yearbook, Liaocheng Urban Master Planning (2014–2030), Statistical Communiques and Annual Work Reports of the National Economic and Social Development, and the official websites of the Liaocheng Statistics Bureau ( http://tjj.liaocheng.gov.cn/), the Liaocheng Culture and Tourism Bureau ( http://wlj.liaocheng.gov.cn/), the Liaocheng Natural Resources and Planning Bureau ( http://zrzyhghj.liaocheng.gov.cn/), and BIGEMAP software. The tourism industry data of Yangzhou were collected from the Yangzhou Statistical Yearbook, Master Planning of Yangzhou Tourism Development (2010–2030), Statistical Communique on National Economic and Social Development, the official websites of the Yangzhou Statistics Bureau ( http://tjj.yangzhou.gov.cn/), the Cultural Radio, Television and Tourism Bureau ( http://wglj.yangzhou.gov.cn/), and BIGEMAP software.
Table 1
Comparison of the major tourism activities along the Beijing-Hangzhou Grand Canal
The data were then verified by Baidu Maps ( https://map.baidu.com/) and processed using the registration and digitization functions of ArcGIS software. By performing these steps, the data for the tourism industries in Beijing, Liaocheng, and Yangzhou in 2010, 2015, and 2019 were collated (Table 2).
Table 2
Tourism industry data for Beijing, Liaocheng, and Yangzhou
Table 3
Influencing factors and data sources
3.2.2 Influencing factors
When determining the factors that affect the tourism industry along the Beijing-Hangzhou Grand Canal, this study considered the characteristics of LCH and factors that influence tourism development in general. Based on relevant studies, eight factors referring to canal resources, the natural environment and socioeconomic development were selected as the influencing factors, and the corresponding data for Beijing, Liaocheng, and Yangzhou at the three time points (2010, 2015, and 2019) were collected (Table 3).
3.3 Data analysis
3.3.1 Nearest neighbor analysis
The spatial distribution pattern was defined by nearest neighbor analysis, which examines the distances between each point and the point closest to it and then compares these to the expected values in the complete random spatial pattern (Long et al., 2018). Generally, the spatial distribution pattern presents one of three modes: dispersed, randomly distributed, or clustered, which can be determined by the Nearest Neighbor Index (NNI). This method was used in this study to examine spatial agglomeration of the tourism industry along the LCH. The calculation formula is:
where di is the distance between each point and its nearest neighbor; n is the number of points; and A is the size of the area under study. When NNI = 1, the points are randomly distributed; when 0<NNI<1, the points are clustered; and when NNI>1, the points are dispersed.
3.3.2 Kernel density estimation
Kernel density estimation is a non-parametric method for measuring the probability density function of variables, which provides a smooth, intuitive expression by transforming discrete points into a continuous density surface (Yu et al., 2016). The kernel density map of the tourism industry was used to show the temporal-spatial changes in LCH tourism in this study. The variables of h and d were adopted as defaults. The calculation formula is:
where 
(x) is the estimated density value at location x; K[ ] is the kernel function; h is the bandwidth; n is the total number of features within the bandwidth; d is the dimension of the data; and (x–xi) is the distance between feature xi and location x.
3.3.3 Standard deviational ellipse
The standard deviational ellipse (SDE) was used to represent the degree of spatial dispersion and orientation of the tourism industry. Its attribute values are the x and y coordinates for the mean center, the standard distances along the long and short axes, and the orientation of the ellipse (Long et al., 2018). Specifically, the mean center denotes the central position of the arithmetic mean of tourism industry. The lengths of the long and short half axes indicate centripetal and directional feature, while the azimuth angle records the clockwise rotation angle from due north to the long axis of the ellipse. The coordinates and standard deviations of the SDE are calculated as follows:
where SDEX and SDEY represent the coordinates of the ellipse; xi and yi are the coordinates for feature i; x and y are the mean centers for the features; n is the total number of features; σ x and σ y are the standard deviations for the x-axis and y-axis respectively; x#x0304;i and ȳi are the differences between the mean center and the coordinates; and θ is the orientation of the ellipse.
3.3.4 Geo-detector
Geodetector is an effective tool for exploring the causes and mechanisms that drive the spatial patterns of geographical elements. This study explored the explanatory power and interactions of different factors influencing the tourism industry distribution along the LCH according to the factor detector and interaction detector in the Geo-detector.
The factor detector in the Geo-detector is used to detect whether a certain geographical factor is the cause of the spatial distribution difference. Its theoretical core is the detection of the consistency of spatial distribution patterns between dependent and independent variables via spatial heterogeneity, to measure the explanatory degree of independent variables to the dependent variables (Wang and Xu, 2017). The Geo-detector uses the power of the determinant (qx) to reflect the spatial correlation by using the following equation:
where N is the number of samples in the study area; Nh is the number of samples in zone (category) h of factor X; σ 2 is the total variance of Y within zone (category) h of factor x; and L is the number of zones (categories) of factor X. The L term 
is the within-sum of variances; and Nσ 2 h=1 is the total sum of variances. The greater the value of qx, the more factor x explains Y, and vice versa.
The interaction detector is another advantage of the Geo-detector over other statistical methods, and it verifies the interactive influences of factors (Wang and Xu, 2017). The aim is to evaluate whether the combination of factors increases or decreases the explanatory power of the dependent variable.
The method of natural breaks in ArcGIS was adopted to divide each influencing factor into three grades based on the distribution law of the data. Next, this study regarded the numbers of tourism businesses within 15 km buffer zones in Beijing, Liaocheng, and Yangzhou as the dependent variables and influencing factors as the independent variables to conduct the Geo-detector analysis.
4 Results
4.1 Spatial-temporal distribution characteristics of the tourism industry
4.1.1 Increase in number
(1) Beijing section
As the capital and must-go historical destination in China, Beijing has substantial appeal to tourists, although its canal-themed tourism has not attracted much attention. To highlight the tourism potential and influence of the canal, this study used the canal as the axis to draw buffer zones within 5 km, 10 km, and 15 km, and the numbers of business points within these buffer zones were counted (Table 4).
Figure 2 (a–c) shows that the buffer zone within 15 km of the canal is the core area for tourism development in Beijing, with over 65% of the businesses. Although the growth rate is lower than the rate outside the buffer zone, canal tourism has unique advantages in terms of geographical location.
(2) Liaocheng section
Liaocheng was an important shipping town in ancient times, but as the canal deteriorated, its shipping gradually declined. Considering the canal conditions and economic development of Liaocheng, buffer zones within 5 km, 10 km,and 15 km were drawn (Fig. 2d–2f), and the numbers of commercial businesses related to the tourism industry in the corresponding zones were counted (Table 4).
Table 4
Changes in the number of businesses within each buffer zone
From 2010 to 2019, tourism businesses in Liaocheng grew rapidly, spreading from a single core to a multi-core distribution mode. The improvement in tourism in the non-canal areas of Liaocheng was much higher than that along the canal; most notably, the tourism industry along the canal has not significantly changed after its designation as a World Heritage in 2014.
(3) Yangzhou section
Yangzhou is an important city and a typical representative of the Beijing-Hangzhou Grand Canal. This study extracted buffer zones within 5 km, 10 km and 15 km and the numbers of tourism businesses were counted (Fig. 2g-2i, Table 4).
The distribution of tourism businesses in Yangzhou presents an obvious multi-group agglomeration pattern, with Hanjiang District as the main group, and Jiangdu District, Gaoyou City (county-level city), and Baoying County as the secondary groups. However, the tourism industry in the buffer zones improved more slowly than in other areas in the following four years, possibly because of tourism saturation and overloading of the canal areas.
4.1.2 Improvement in agglomeration degree
(1) Beijing section
The results of the NNIs revealed that businesses within 15 km of the canal were organized according to clustered patterns in 2010, 2015, and 2019 (Table 5). The changes in the NNI indicate that the overall agglomeration degree of businesses weakened with the expansion of spatial distribution after 2015, although it had strengthened significantly in the previous five years. Moreover, the patterns of each type of tourism business in the different years are clustered, with lower indices than those of the overall businesses. For example, the NNIs of tourism landscape, tourism shopping, accommodation facilities, leisure catering, and entertainment in 2010 varied from 0.3 to 0.6, and all showed spatial agglomeration. The degree of agglomeration of each type improved in the following five years.
(2) Liaocheng section
The NNIs of Liaocheng tourism businesses within the 15 km buffer zone were 0.218, 0.171, and 0.200 in 2010, 2015, and 2019, respectively (Table 5). Although the degree of agglomeration increased at first and then decreased, the businesses were always clustered. Specifically, the NNIs in 2010 showed highly clustered characteristics. In 2015, the degrees of agglomeration of tourism landscape and accommodation facilities were promoted, and the other three were reduced, especially the NNI of tourism shopping which increased to 0.499. The NNIs of all the business subdivisions showed upward trends in 2019, resulting in attenuated aggregation degrees.
(3) Yangzhou section
The NNIs of Yangzhou within the 15 km buffer zone in 2010, 2015, and 2019 were 0.134, 0.166, and 0.152, respectively (Table 5). Overall, the businesses at each time point showed clustered distribution patterns, but the degree of agglomeration tended to weaken. From 2010 to 2015, the number of businesses 15 km away increased rapidly, reducing the degree of agglomeration. On this basis, the NNIs of accommodation facilities and tourism landscape continued to grow in 2019, and the tourism landscape ranked first. However, leisure catering and entertainment returned to their 2010 levels.
4.1.3 Changes in spatial structure
(1) Beijing section
The results of kernel density estimation showed that tourism businesses in Beijing from 2010 to 2019 presented a circular distribution around the high-value areas. With the continuous expansion of high-value areas, the overall distribution of businesses evolved from a small-scale mononuclear pattern to a multi-core pattern (Fig. 3a-3c). Equipped with perfect tourism support services, the Shichahai and Tonghui River basins in the Dongcheng and Xicheng districts were the most frequently visited destinations of the urban tourism industry. In 2019, the recreational construction along the canal in the section west of the Sihui Bridge was completed. With the development of the city subsidiary-center, Tongzhou District promoted canal culture and reshaped its landscapes by building Canal Cultural Square, Canal Olympic Park, and Tongzhou Grand Canal Forest Park; improving transportation and basic infrastructure; and creating golf clubs, resorts, and other leisure businesses.
A directional distribution analysis of the buffer zone within 5 km demonstrated that the central point of the ellipse in each of the three years remained almost unchanged, but the increase in the ellipse area indicated that the tourism industry has been expanding continually (Table 6). The expansion direction of businesses was horizontal from 2010 to 2015, and both the long axis (X-axis) and the short axis (Y-axis) increased simultaneously over the last five years.
(2) Liaocheng section
The distribution of tourism businesses in Liaocheng changed from single-core agglomeration to multi-core agglomeration, and then to multi-core circular agglomeration from 2010 to 2019 (Fig. 3d-3f). The high-density areas mainly involved Dongchangfu district and Linqing city within the 15 km buffer zone along the Beijing-Hangzhou Grand Canal. As the center of Liaocheng, Dongchangfu District relies on the Water Park, Dongchang Lake, and Nanhu Wetland Park to create an entertainment area, which conforms to Liaocheng's urban positioning of the Ancient Watery City and promotes shopping, accommodation, and other commercial activities to form the main business center. Linqing's commercial vitality is obviously weaker than that of the Dongchang District. Moreover, in Yanggu County, south of Liaocheng City, two small, high-value clusters of tourism industries have appeared successively since 2015, in the towns Qizi and Zhangqiu.
Table 5
NNIs of the tourism industry types in Beijing, Liaocheng, and Yangzhou
The directional distribution of tourism businesses along the canal within the 5 km buffer zone was characterized by a continuous westward and northward movement in the past decade (Fig. 3d-f, Table 6). The ellipse area expanded from 2010 to 2015, and the X-axis and Y-axis both showed substantial growth. However, the two axes shortened in the next four years, indicating the emergence of industrial concentration based on horizontal and vertical diffusion. Overall, the line of the high-value clusters located in Dongchang District and Linqing City was the developmental direction of the tourism businesses.
(3) Yangzhou section
The high-value areas of businesses were mainly distributed in the junction of Guangling District and Hanjiang District, which were the ancient city and downtown of Yangzhou, respectively (Fig. 3g-i). The buffer zone within 15 km of the canal covered all high-value agglomeration areas of Yangzhou's tourism businesses, indicating that the canal was the main tourist attraction of Yangzhou. The area from south of the Slender Westlake and Benyimen Pier, west of the ancient Canal to Yinchaohe Park and Yangzhou Museum, was the main tourism industrial agglomeration area. It includes scenic spots, such as Dongguan Street, the Slender Westlake, and He Yuan, in addition to heritages listed as World Cultural Heritage sites, such as Tianning Temple, Chongning Temple, and Ge Garden. In addition, the tourism high-value clusters in Baoying County and Gaoyou City were on the right bank of the canal. The businesses are mainly distributed in the center of Baoying County, with relatively perfect marketing service facilities.
The directional distribution analysis of the buffer zone within 5 km showed that the central point of the ellipse moved to the north and east during the past decade (Table 6). Specifically, it was located west of the Slender Westlake in 2010, and then moved to the west of Shaobo Lake in 2019. Simultaneously, the ellipse area continued to expand, which mainly resulted from the almost doubling of the long axis (Fig. 3g-3i).
4.2 Factors influencing the tourism industry distribution
4.2.1 Explanatory power of influencing factors
The explanatory power of socioeconomic development is stronger than those of canal heritage and the natural environment (Table 7). Regarding the basic information of the canal, X1 (number of heritage sites), X2 (canal water quality) and X3 (channel length) are highly associated with the distribution of tourism businesses along Beijing-Hangzhou Grand Canal, with high q values above 0.75. The q values of air quality (X4) and urban green coverage rate (X5) are the lowest among all factors, at 0.312 and 0.169, respectively, indicating that the natural environment only slightly influences the tourism industry distribution. By contrast, factors belonging to socioeconomic development possess high q values. Most notably, the value of X6 (number of tourists) reached 0.902, which ranked first among all factors. Two of the three factors in canal heritage and socioeconomic development have equal q values, while the others (X1 and X6) show different characteristics. X1 affects tourism business distribution slightly with a q value of 0.211 and X6 plays a significant role in the temporal-spatial pattern of businesses.
4.2.2 Interactions of influencing factors
Table 8 shows the interaction q-statistic values of the influential factors. Generally, the explanatory power of the interaction between any two factors is higher than that of a single factor. Their interactions represent significant nonlinear enhancement, indicating that the combined effect of any two factors increases the explanatory power of the tourism industry distribution.
Fig. 3
Directional distributions and kernel densities of businesses in Beijing, Liaocheng, and Yangzhou
Table 6
Attributes of standard deviational ellipses of the tourism industries in Beijing, Liaocheng, and Yangzhou
From the explanatory power of X1 (number of heritage sites) alone, the addition of factors related to canal heritage (X2 and X3) or socioeconomic development (X6, X7, and X8) both enhance the explanatory power of tourism businesses along the Beijing-Hangzhou Grand Canal. However, the addition of air quality (X4) or urban green coverage (X5) only slightly enhanced the explanatory power, indicating that the natural environment is not very attractive to visitors and has no marketing advantage. Notably, although X4 has a low q value of 0.312, when it interacts with factors other than X1 and X5, the interactive q-statistic values exceed 0.94. A similar situation occurs with X5, which shows a significant nonlinear enhancement of the explanatory power above 0.9 when it interacts with X6 (number of tourists) or X7 (GDP).
5 Discussion
The spatial evolution of the tourism industries in cities along the Beijing-Hangzhou Grand Canal is similar to that of cities along other LCH, and is an inevitable result of the spatial expansion along the linear space. In terms of spatial disparity, some studies have made comparisons from the perspective of the LCH as a whole (Caton and Santos, 2007; Božić and Tomić, 2016; Zhang et al., 2020). However, considering the complicated microcosmic environments and rapid business expansion in the destinations (Briedenhann and Wickens, 2004), it is worth comparing the tourism industry distribution in the different cities along an LCH. Businesses in Beijing, Liaocheng, and Yangzhou have presented distinctly different temporal-spatial structures, although the number and degree of agglomeration increased substantially from 2010 to 2019. Basically, the tourism industrial agglomerations along the canal are attributed to geographical proximity in these three cities, indicating that the canal remains an important impetus for regional economic and cultural development. The tendency for the direction of commercial expansion to be consistent with the flow direction of the canal appears to be the same in Beijing, Liaocheng, and Yangzhou (Fig. 3). While this tendency does not mean that it is the only agglomeration mode, it still provides guides for urban tourism marketing by capturing the supply-oriented preferences. Notably, the distributions of tourism hotspots within cities require the consideration of additional extensive factors (Deng et al., 2014).
Table 7
Results of the factor detector
Table 8
Results of the interaction detector
Although the outstanding universal value of the Beijing-Hangzhou Grand Canal is recognized globally, a substantial disparity exists in tourism development along the canal (Li and Zou, 2021). Based on the temporal-spatial distribution characteristics, several key observations on the influencing factors of tourism industry distribution along the Beijing-Hangzhou Grand Canal are made. First, most studies have verified the strong correlation between heritage resources and the tourism industry along LCH (Božić and Berić,2013; Ren, 2017; Gravari-Barbas, 2018; Li, 2019), which further demonstrates that the quality of resources (X2 and X3) is more important than the quantity of sites (X1). In the process of transforming heritage resources into tourism products, high-quality derivative products are closely bound to a wide range of tourist markets and high satisfaction, which significantly relies on the protected status of the heritage resources themselves (Vana and Malaescu, 2016). Therefore, LCH destinations should pay more attention to the marketing and publicity of high-quality resources.
Second, this study found that the natural environment plays only a small role in forming the temporal-spatial pattern of the tourism industry along the canal, which is in accordance with research conducted by Deng et al. (2014) and Zhang et al. (2020). Considering the geographical location of the Beijing-Hangzhou Grand Canal, only minor differences in the natural environment are observed among the regions (Zhang et al., 2019), indicating that the ecosystem has no obvious market advantage for the cities along the canal. However, this finding is not applicable to other LCH, such as the Silk Road or the Norwegian St. Olav Pilgrim Routes.
Third, the influencing factors related to socioeconomic development are the most active, regardless of the tourism market, economic development, or transportation, all of which are crucial topics for discussion. Undoubtedly, tourism industries are deeply rooted in socioeconomic backgrounds (Ren, 2017). Tourism is a complex, comprehensive industry that involves various economic and social aspects (Briedenhann and Wickens, 2004). The higher the level of socioeconomic development, the more active the market atmosphere (Du and Liu, 2011), which stimulates entrepreneurs' enthusiasm for tourism investment and provides adequate support from the community (Murray and Graham, 1997). The temporal-spatial layout of tourism businesses is not only limited by the aforementioned factors; therefore, in-depth discussions on additional influential factors in different LCH are necessary.
6 Conclusions and implications
6.1 Conclusions
Using Beijing, Liaocheng, and Yangzhou as study sites, this study explored the temporal-spatial distribution characteristics and influencing factors along the Beijing-Hangzhou Grand Canal. The results may be applicable to other LCH that are experiencing the challenge of heritage reuse and aim to develop the tourism industry. The analysis results show a continuous increase in the quantity and agglomeration degree of travel-related businesses from 2010 to 2019 in either Beijing, Liaocheng, or Yangzhou. Furthermore, the direction of industrial expansion in these three cities is consistent with the flow direction of the canal in view of the spatial structure. Moreover, the distribution characteristics of tourism businesses are found to have formed under the effects of multiple influencing factors. The explanatory powers of factors related to socioeconomic development and canal resources are stronger than those related to the natural environment.
6.2 Implications
6.2.1 Theoretical implications
This study analyzed the spatial evolution of the tourism industry along the Beijing-Hangzhou Grand Canal by comparing individual cities from a dynamic perspective, which enriches the theoretical research on LCH management and recreational reuse. Previous studies have assessed tourism utilization from the perspective of LCH as a whole, but few have compared the complicated microcosmic environments in areas along an LCH, especially at the city level (Zhang et al., 2023b). Underpinned by multidisciplinary theories and methods, this study conducted a comparative analysis of the three cities of Beijing, Liaocheng, and Yangzhou. Unlike other studies that chose one famous section with high-level tourism development, this study considered the different combinations of heritage function and tourism development to reveal the laws governing the LCH tourism industry under various special settings. Our findings on the tourism industry of the Beijing-Hangzhou Grand Canal, especially the temporal-spatial distribution laws in the three cities, enhance the research contents of tourism geography and cultural heritage management, and also improve academic research in related fields.
The identification of influencing factors is another highlight of this study. By comparing the temporal-spatial distributions of the tourism industries in three cities, this study tried to clarify the internal logic of the evolutionary mechanism and spatial differences. Although tourism development is progressing in the Beijing-Hangzhou Grand Canal, spatial disparity remains a problem that must be solved. The findings on the influencing factors provide ideas for revealing the business layouts of LCH or similar tourism resources, and are essential basic research for location theory and tourism marketing. The discussions about the explanatory powers of the different factors underscore the complex situations faced by LCH, which is conducive to dialectically treating LCH tourism and striking a balance between heritage protection and tourism development. By deepening our understanding of this heritage type, this study provides theoretical guidance for existing large-scale heritage management and policy adjustments. This also has numerous implications for the tourism marketing and sustainable optimization of the Beijing-Hangzhou Grand Canal. In addition to combining the influencing factors, further research should aim to optimize the spatial layout of travel-related businesses and formulate marketing strategies by adjusting the explanatory powers of the various factors.
6.2.2 Practical implications
Based on the above analysis of the temporal-spatial distribution of the Beijing-Hangzhou Grand Canal and tourism-related businesses, different developmental countermeasures should be adopted for the different types of cities along the canal. Taking Beijing, Liaocheng, and Yangzhou as examples, the following management responses are proposed for the improvement of tourism in the Beijing-Hangzhou Grand Canal.
For the Beijing section of the canal, it is necessary to repair or restore the canal ruins, fully excavate the canal culture, and create a unique canal image, so that the canal tourism can stand out in the fierce market competition. At the same time, efforts should be made to enhance the connections between the canal and other heritage and scenic spots, strengthen the improvements of the surrounding environment, and protect the ecological environment of the canal.
For Liaocheng section, with its damaged heritage sites and limited tourism businesses, strengthening the management of the canal and heritage is the basis and key for maintaining the canal culture. Furthermore, speeding up the environmental transformation, implementing ecological greening projects, and promoting the universal education of canal culture are all feasible solutions. In the spatial layout of the tourism industry, it is necessary to form one or several strong tourism poles to drive the rapid development of the entire canal tourism industry.
The canal tourism in Yangzhou section has a good demonstration role, benefiting from the specific regional culture, economic environment and policy measures. To achieve further optimization, the Yangzhou section should have a deep understanding of the demand of the tourist source market, formulate adaptive development plans according to the requirements of the local resource level and protection and development. At the same time, it is necessary to balance the distribution of stakeholder interests, guide the healthy competition of tourism enterprises, and avoid the loss of cultural connotation and the destruction of cultural heritage that can be caused by excessive commercialization.
