INTRODUCTION

The magnitude of the crisis caused by the COVID-19 pandemic is unprecedented. Almost all countries and all social spheres have been affected in one way or another (Casero-Ripollés, 2020), although tourism has been one of the economic sectors that have suffered the most from its consequences. Not in vain, tourism depends largely on mobility and transportation, which were paralyzed due to the restrictions imposed by governments around the world to curb the incidence of the virus (Yang, Zhang, & Chen, 2020).

It must also be taken into account that tourists are often reluctant to travel to a destination where they could suffer any danger (Fuchs & Reichel, 2011), although this fear increases when the risk has to do with the possibility of suffering a disease (Araña & León, 2008). In fact, throughout the 21^st century, there have been many examples of health crises that have had a negative impact on tourism.

Just twenty years ago, SARS caused a health alarm that particularly affected tourists who intended to travel to Asian countries, which also had a huge impact on tourism employment (Mckercher & Chon, 2004). Likewise, both influenza A (H1N1) in 2009 and the coronavirus (MERS.Cov) in 2012 significantly affected the tourism sector, especially in Asia (Joo, Henry, Lee, Berro, & Maskery, 2019; Lee, Song, Bendle, Kim, & Han, 2012).

Now, the COVID-19 pandemic has caused widespread losses in the tourism sector around the world. Between January and August 2020, international traveler arrivals suffered a 70% decrease compared to the same period in 2019 (UNWTO, 2020). The economic losses due to this fall exceeded 700,000 million euros.

On the other hand, this new pandemic has caused a crisis that has affected all social structures. The field of corporate communication has also been surprised by a crisis that could not be foreseen and that will have to be analyzed to obtain conclusions that help better communication management of this type of situation (Xifra, 2020).

As far as tourism is concerned, the DMOs (Destination Marketing Offices) saw the need to seek promotional alternatives and convey safety and trust among potential tourists and visitors (Oliveira & Huertas-Roig, 2019; Paniagua-Rojano & Pastor-Marín, 2021). In those first months, in which travel was prohibited, it was obviously about trying to keep alive the interest of the demand for traveling to destinations once the situation returned to normal (Huertas, Oliveira, & Girotto, 2020).

In this sense, social media can be of great importance when it comes to managing crises (Cristòfol, De-San-Eugenio-Vela, & Paniagua-Rojano, 2020; Pastor-Marín, González, & Paniagua-Rojano, 2021). For this reason, there are more and more studies focused on the use of social media to manage communication in times of crisis (Landau, 2011; Graham et al., 2015; Jin et al., 2011; Lin et al., 2016; Liu and Faustino, 2014; Utz et al., 2013; Van der Meer and Verhoeven, 2013, 2014; Zhu, et al., 2017) (Graham, Avery, & Park, 2015; Jin, Liu, & Austin, 2011; Landau, 2011; Lin, Spencer, Sellnow, & Lachlan, 2016; Liu & Faustino, 2014; Utz, Schultz, & Glocka, 2013; Zhu, Anagondahalli, & Zhang, 2017).

In particular, social media have been fundamental in situations of uncertainty caused by pandemics or health crises (Chew & Eysenbach, 2010; Gui, Kou, Pine, & Chen, 2017; Veil, Buehner, & Palenchar, 2011). Among the most recent studies, the research carried out byHuertas et al. (2020) stands out, which highlighted the lack of true crisis management by the national offices of Spain and Italy, especially during the first months of the pandemic.

Also, Paniagua-Rojano and Pastor-Marín et al. (2021), when studying the communication carried out by some of the most important urban destinations in the world on Facebook and Twitter in the first wave of the pandemic, stressed that, despite the increase in information on services, safety, and health, the absence of a concrete plan to use social media as management instruments in the crisis was evident.

OBJECTIVES

This study aims to analyze whether some of the most representative destinations in Spain and Portugal have a real crisis communication strategy on Twitter and how they have used this tool during the COVID-19 pandemic. Based on this main objective, we have studied how the tourist destinations that are part of this research use Twitter, what type of publications facilitate interaction with the public, and whether or not these have been used by those responsible for marketing the destinations during the pandemic. In any case, the initial hypothesis of this research is that the DMOs have not specially engaged in dialogue with travelers during the crisis and have continued to use Twitter solely as a means of information.

To achieve the stated objectives, all the publications of the official Twitter profiles of twelve destinations in Spain and Portugal, made from November 21^st to December 21^st, 2020, have been analyzed. This period -before the Christmas holidays- allows for studying communication on social media on dates that are traditionally very significant for the tourism sector. It is also taken into account that this month coincided with the third coronavirus wave, which had a high impact on both countries. In total, 1,382 tweets have been analyzed.

The sample is made up of a total of six Spanish and six Portuguese tourist destinations, which correspond to the geographical areas used by the market research company Nielsen, taking into account a series of more or less similar characteristics.

The sample includes five types of destinations for each country: region or autonomous community (Andalusia, Algarve), large coastal destination (Balearic Islands, Azores), large municipality (Madrid, Lisbon), medium-sized city (Málaga, Porto), and inland destination (Castilla y León, Central Portugal), to which the national Tourism accounts of Spain and Portugal (Turespaña and Visit Portugal) are added.

Thus, this research analyzes the content of the accounts @viveandalucia, @visitalgarve_es, @TurismeBalears, @visitingazores, @Visita_Madrid, @TurismodeLisboa, @turismodemalaga, @visitporto, @CyLesvida, @CentroPortugal, @Turespana_, and @visitaportugal.

METHODOLOGY

For the study of the publications made by the destinations in the indicated period, the Fanpage Karma tool (www.fanpagekarma.com) has been used, which allows an adequate analysis of the activity of the profiles and the interaction of the public with the tweets. This tool has been used in similar research works, which have shown its usefulness for measuring aspects such as reactions to certain content (García, 2015; Olabe and Márquez, 2019; ) (Pastor-Marín & Paniagua-Rojano, 2020; Pastor-Marín et al., 2021).

The use of Fanpage Karma has made it possible to monitor relevant data such as how many followers each of the profiles has, the performance index of the publications, or what type of reactions it provokes among the public. It has also allowed the analysis of the used hashtags or the potential of influence, for which the number of profiles followed by each of the accounts has also been taken into account.

Once the data has been monitored, the methodology focuses on content analysis, which is considered an ideal technique for studying the posts made on Twitter (Castelló et al., 2014). In this phase, we analyze which tourist aspects are the most mentioned or communicated by the DMOs and if they correspond to those that generate the most reactions, to also see what space the destinations dedicated to information related to the crisis generated by the pandemic.

For the content analysis, the publications are divided into the following topics or values: tangible heritage, landscape, agenda, climate, nature, services, leisure, institutional messages, intangible heritage, sports, business, safety, health, and technology.

RESULTS AND DISCUSSION

On average, the analyzed destinations have 48,496 followers on Twitter, although the differences between them are notable (Figure 1). The official tourism account of Portugal leads the classification in this area, with 110,813 followers, followed by Andalusia (109,444), Castilla y León (106,307), Lisbon (78,893), Madrid (74,245), the Balearic Islands (34,706), and Malaga. (30,499). The destinations with the fewest followers are Algarve (2,671), Porto (6,045), Azores (6,462), Turismo do Centro (10,817), and Spain (11,060).

It can thus be seen that the largest destinations or even those that receive a greater number of travelers do not have to be the ones with the most followers on Twitter. There is also no obvious pattern between the different types of destinations so that one national account (Portugal) is the one with the highest number of followers while the other (Spain), despite representing one of the countries with the highest number of travelers in the world, is the one with the fewest followers.

Although the use of Twitter by tourist destinations is still under discussion, the results of this study conclude that this social media platform, one of the most used on the Internet (Akehurst, 2009; Kirilenko & Stepchenkova, 2014; Krikorian, 2013; Thelwall, Buckey, & Paltoglou, 2011), is of great importance for the tourism promotion of the great destinations of Spain and Portugal. Not surprisingly, all the destinations analyzed have a notable presence on Twitter and a significant potential for influence.

On the other hand, in all destinations, the number of followers is greater than the number of profiles they follow. On average, each destination follows 1,258 accounts. As in the previous section, Portugal also leads this section, with 4,730 accounts followed. Next are Andalusia (3,680), the Balearic Islands (2,270), Malaga (1,031), Madrid (818), Spain (445), Azores (427), Porto (426), Castilla y León (407), Turismo do Centro (370), Algarve (309), and Lisbon (189).

The influence potential of a Twitter account can be calculated by dividing the number of followers by the number of followed accounts (Pastor et al., 2021). In this case, the average is 38.5. At the head is Lisbon (417.4), followed by Castilla y León (261.1), Madrid (90.7), Andalusia (29.7), Málaga (29.5), Turismo do Centro (29, 2), Turespaña (24.8), Portugal (23.4), the Balearic Islands (15.2), the Azores (15.1), Porto (14.1), and Algarve (8.6).

The activity of destinations on Twitter is also uneven. On average, in this period each destination published 115 tweets. The destination with the highest number of publications was Castilla y León (485), followed by Portugal (232), Spain (135), Andalusia (123), and Madrid (117). For their part, the destinations that published the least were Porto (61), Lisbon (58), the Balearic Islands (54), Malaga (42), Algarve (18), Turismo do Centro (28), and Azores (29), which reflects that it is not the destinations that publish the most that have a higher level of influence.

The average number of daily posts is therefore 4.1. Once again, Castilla y León heads the ranking (17), followed by Portugal (8.3), Spain (4.8), Andalusia (4.4), Madrid (4.2), Porto (2.2), Balearic Islands (2.1), Lisbon (2.1), Malaga (1.4), Turismo do Centro (1), Azores (1), and Algarve (0.6).

The destinations, however, do not maintain the same level of activity throughout the week. Fewer posts are usually made on weekends. In fact, Sunday is the day with the fewest number of tweets (107), closely followed by Saturday (114). During the week, the average remains stable between 229 on Mondays and 256 on Fridays.

Source: Own elaboration.

In the field of social media, one of the most outstanding aspects is always engagement, which allows us to analyze the results of a communication strategy. This commitment is closely linked to interaction with users. In this research, the average number of reactions, comments, and tweets shared stands at 3,421 (Figure 2), although the differences between some destinations and others are especially significant: Castilla y León (23,665) heads the ranking, followed by Lisbon (4,392), Portugal (3,947), Andalusia (3,188), Madrid (2,900), the Balearic Islands (845), Azores (793), Spain (519), Algarve (360), Malaga (225), Turismo do Centro (149), and Porto (77).

In the analyzed period, the average of 'likes' by destination reaches the figure of 2,946. However, as in the rest of the parameters studied, numerous differences are observed between the different profiles. Once again, Castilla y León (20,464) is at the top, followed by Lisbon (3,963), Portugal (3,419), Andalusia (2,605), Madrid (2,400), the Balearic Islands (709), the Azores (701), Spain (401), Algarve (313), Malaga (183), Turismo do Centro (134), and Porto (68).

In this way, each of the publications receives an average of 19 likes. In this variable, @VisitLisboa is the profile in which the publications achieve a higher performance index, with an average of 68.3 likes for each tweet. Following are Castilla y León (42.9), Azores (24.1), Andalusia (21), Algarve (17.3), Madrid (17), Portugal (14), the Balearic Islands (12.2), Malaga (4.9), Turismo do Centro (4.7), Spain (3), and Porto (0.9).

Source: Own elaboration.

On average, each posted tweet is retweeted 2.9 times. At the head is Lisbon (7.9), followed by Castilla y León (6.7), Andalusia (4.7), Madrid (3.7), Azores (3.1), Algarve (2.6), the Balearic Islands (2.3), Portugal (2.1), Malaga (1.1), Spain (0.8), Turismo do Centro (0.5), and Porto (0.1).

On the other hand, the participation of destinations in conversations with other users is scarce. The average is 18.7%, but up to four destinations do not participate in this type of dialogue (Málaga, Turismo do Centro, Algarve, and Andalusia). Those that participate the most are the Balearic Islands (50%), the Azores (50%), Porto (36.5%), Lisbon (33.3%), and Madrid (29%). Spain does 13.7%, Portugal 8%, and Castilla y León 4.7%.

With so little participation, logically a very low rate of interaction is also observed, with an average of 0.12%. None of the destinations reaches even 1%. Algarve (0.73%) leads the ranking, although closely followed by the Azores (0.43%), Lisbon (0.099%), Turismo do Centro (0.051%), Castilla y León (0.047%), the Balearic Islands (0.042%), Turespaña (0.036%), Andalusia (0.023%), Porto (0.022%), Madrid (0.021%), Malaga (0.020%), and Portugal (0.014%).

Another element that helps the study of interactivity on Twitter is that of retweets. In this case, each destination replicates an average of 432 posts (Figure 3). Castilla y León is the destination that retweets the most. Far behind are Andalusia (583), Portugal (528), Madrid (484), Lisbon (429), the Balearic Islands (136), Spain (118), the Azores (92), the Algarve (47), Malaga (42), Turismo do Centro (15), and Porto (9).

Source: Own elaboration.

Image is a substantial part of the publications made in this period. Almost a third of the tweets -30.5%- include a photograph. It is in this area, however, where the greatest differences between the profiles are found. While the Azores accompany their tweets with images in 96.6% of cases, neither Lisbon nor Porto use their publications solely with photographs.

In the rest of the destinations, the differences are also great: Algarve (72%) is followed by Castilla y León (66.9%), the Balearic Islands (40%), Madrid (21.4%), Malaga (19%), Andalusia (14.5%), Turismo do Centro (14.3%), Spain (13.3%), and Portugal (8.2%).

In second place are the tweets that, besides the image, include a link (27.5%). Lisbon is the destination that uses this resource the most (98.3%), followed by Andalusia (71.8%), Malaga (42.9%), the Balearic Islands (33.3%), Castilla y León (27.3%), Spain (25.2%), Portugal (18.5%), Turismo do Centro (10.7%), Porto (1.6%), and Madrid (0.9%). Neither Algarve nor the Azores uses this resource.

Destinations use only text in 21.7% of their tweets. Madrid (61.5%) uses this type of publication the most, followed by Portugal (52.2%), Spain (48.9%), Malaga (31%), the Balearic Islands (25%), Porto (24, 6%), Andalusia (12.1%), the Azores (3.4%), Lisbon (1.7%), and Castilla y León (0.2%). Neither Algarve nor Turismo do Centro publish text-only tweets.

Finally, 20.2% of tweets use a link to other information. Turismo do Centro uses links in 75% of its tweets, followed by Porto (73.8%). Next are Algarve (27.8%), Portugal (21.1%), Madrid (16.2%), Spain (12.6%), Malaga (7.1%), Castilla y León (5 .7%), the Balearic Islands (1.7%), and Andalusia (1.6%). Neither Azores nor Lisbon publishes these types of tweets.

76.8% of the tweets published between November and December 2020 -that is, 10,062 tweets- were new content, produced by the DMOs. The rest of the publications were content shared with other accounts (272 tweets, 19.7% of the total) or responses (48 tweets, 3.5%), which once again highlights the low level of interaction with the public. The destinations with the highest percentage of new content are Turismo de Centro and Andalusia, with 100%, followed by Castilla y León (99.4%) and Lisbon (98.3%).

The destinations that respond the most are Porto (18%), Madrid (16.2%), Spain (5.9%), the Azores (3.4%), the Balearic Islands (3.3%), and Portugal (2.2%). In the analyzed period, Algarve, Castilla y León, Malaga, Turismo do Centro, Lisbon, and Andalusia did not respond.

On Twitter, the use of hashtags is considered an excellent tool to encourage dialogue and interaction with the public, while facilitating the search for publications (Estebaranz & Ramilo, 2013). In this case, most of the hashtags appeal to the destinations' own tourism brands: #portugal (103), #viveandalucia (90), #academiaviveandalucia (44), #palencia (44), #burgos (43), #zamora (40), #visitamadrid (36), #salamanca (34), #segovia (33), #islasbalearessostenibles (30), #açores (24), #azores (24), and #malagaciudadgenial (21), although it also stands out the use by the official Portuguese account of the hashtag #staysafe (57), which refers to the need to maintain hygienic-sanitary measures to defeat the pandemic.

Similarly, once articles, adverbs, and conjunctions have been ruled out, the words most used by tourist destinations in their publications refer to the names of their best-known territories or tourist elements: León (104), Castilla (92), Andalusia (38), Madrid (34), Portugal (29), Spain (23), Porto (23), or island (16). Furthermore, emoticons used as a symbol of geographical location (53) or photography (52) are increasingly used by destinations.

As for the topics covered in the tweets, it is noteworthy that, despite the pandemic, health information (2.4%) or tweets related to safety (1.7%) are practically residual. However, compared to other previous research (Pastor and Paniagua, 2020), it is evident that some destinations significantly increased the number of tweets dedicated to nature or landscape.

In a pandemic caused by a respiratory virus, these contents have been linked to the segments that could contribute the most to the recovery of tourism by taking place outdoors, which, at least implicitly, could be understood as a strategic way to address the crisis. Thus, although the differences between destinations are important, almost all the DMOs increased the number of tweets on these topics, and Algarve, for example, dedicated 61.1% of its publications to landscapes.

In any case, during the analyzed period, tweets about tangible heritage elements were the majority (20.6% of the total). Once again, significant differences are observed between tourist destinations. Andalusia is the destination that dedicates the most tweets to tangible heritage (27.7% of all its publications). On the other hand, the profile of the Spanish tourist office only dedicates 8.1% of its tweets to tangible heritage.

Next were the publications referring to landscape (16.9%), services (12.4%), nature (11.2%), intangible heritage (10.6%), institutional messages (6.7%), calendar (6.5%), shopping (4.3%), and sports (3.4%). In the last place are tweets related to technology (1.5%) and climate (0.5%).

In any case, most of the tweets focus on issues related to tangible or intangible heritage, landscape, nature, or services. Thus, Algarve, Azores, Turismo do Centro, and Lisbon do not publish institutional messages or topics related to sports; Algarve, Andalusia, and Azores do not publish anything about the calendar; and the only Portuguese destination that publishes about shopping is Lisbon.

Finally, to study which contents are the ones that most arouse the interest of the public, the reactions ('likes', retweets, and comments) of the different publications were analyzed. The largest number of reactions always occurs in those tweets dedicated to the tangible heritage of destinations, which is usually linked to those monumental or artistic elements that form a substantial part of the brand image. Furthermore, among the tweets with the most reactions, none has to do with the pandemic or with issues related to health safety.

Of the sixty tweets with the highest number of reactions, a total of 26 are dedicated to the tangible heritage, which represents 43.3% of the total. In the second place, in terms of the number of reactions, are the publications dedicated to landscapes (15), which represent 25%; followed by tweets featuring elements of intangible heritage (8), 13.3%. The classification is completed with publications on nature (4), which represent 6.6%; and those dedicated to shopping (3) and institutional messages (3), which represent 5%.

In absolute terms, the most 'successful' tweet, in terms of reactions, is one published by Castilla y León, with a photograph of the Burgos Cathedral, a World Heritage Site (350 'likes'), followed by another tweet with a panoramic photo of this same city (331 'likes).

CONCLUSIONS

The research has shown that the destinations analyzed in both Spain and Portugal have not used all the potential that Twitter offers as a means of information and, in fact, the DMOs have used this social media platform during the COVID-19 pandemic as one more way of information on its tourist attractions and, of course, not as a tool for interaction and dialogue with the public.

Likewise, the results show that the crisis generated by the pandemic has not caused those responsible for tourist destinations to modify the content included on Twitter. In fact, even at a time as sensitive for tourism as the third wave of COVID-19, which changed the plans of many tourists when organizing their Christmas holidays, tourist destinations in Spain and Portugal continued to dedicate their tweets to the elements of tangible heritage and only the Portuguese national account (Visit Portugal) accompanied these contents with messages inviting potential tourists to stay home.

Of course, although not unanimously, the destinations increased the number of tweets referring to nature and landscape, linking these publications to the opportunity to enjoy outdoor spaces. However, along the same lines as previous research (Antoniadis, Vrana, & Zafiropoulos, 2014; Pastor-Marín & Paniagua-Rojano, 2020; Savin, 2013), it is found that most DMOs in Spain and Portugal have continued to use Twitter solely as an information tool rather than of dialogue and interaction.

On the other hand, the existence of two different tourism management models has been evidenced. Thus, in Portugal, the accounts with the greatest capacity for influence and the greatest number of followers are those that correspond to the official profile of the country (Visit Portugal), as well as that of the capital city (Visit Lisboa). On the contrary, in Spain, the accounts of the autonomous communities and the cities have a greater weight than the official account of the national tourism body (Turespaña).

Likewise, it has been detected that, although there is room for growth, there has been notable progress in the use of images to reinforce the message of DMOs on Twitter. Thus, most posts use some type of image. In general, the major destinations (Azores, Algarve, Castilla y León, and the Balearic Islands) are the ones that make the most use of this resource, although in absolute terms it is the official Lisbon account that heads the classification, including images and links in almost all of its posts.

In this way, the value of photography is reaffirmed as a means of reinforcing the tourist image, asWhite (2010) had already indicated, in one of the pioneering studies on the use of photography in social media in which he had indicated that images have a great capacity to influence the decisions of future tourists when it comes to organizing their trips.

It is also evident that, although the use of photographs continues to be linked to the promotion of the great tourist attractions of destinations, the images that appeal to the feelings of tourists or have an important symbolic load are the ones that are most liked by tourists, as MMíguez and Huertas-Roig (2015) andPastor-Marín and Paniagua-Rojano (2020) had already pointed out.

Similarly, progress has also been made in the use of hashtags, which, asEstebaranz and Ramilo (2013) explained, have a fundamental role in encouraging dialogue and interaction. In this sense, the DMOs have used these resources correctly since the majority of the words chosen for the hashtags are related to the distinctive elements of each of the destinations or about the need to stay home during the pandemic to enjoy these resources again in the future.

However, despite these advances, it has become clear that interaction with the public is very low.

Destinations do not respond, they do not participate in conversations, and they do not offer solutions to the doubts raised by travelers. To this end, it would be convenient for the DMOs to strengthen their communication strategies, promoting dialogue to achieve greater user participation.

On the other hand, we agree withVijaykumar, Jin, and Nowak (2015) on the importance of using social media to report on relevant aspects of crisis situations caused by pandemics or health crises and, in fact, during the analyzed period, destinations used their official profiles to transmit some type of relevant message about health or safety information related to COVID-19.

In any case, the results show that, despite the seriousness of the crisis, there has been no real communication strategy. For this reason, we conclude that the DMOs should implement a strategic plan that takes full advantage of the potential that social media have to deal with other crises that may arise.

On the other hand, the tweets with the highest number of reactions are those that focus on tangible heritage, while the least successful publications are those that have to do with safety, health, calendar, or technology, that is, those messages that emphasize information rather than emotion or brand image creation.

In this sense, moreover, no notable differences in behavior are observed between Spanish and Portuguese destinations or between the different types of territory (country, region, medium-sized city, large coastal destination, or capital city). All of them should continue to make an effort to increase their ability to interact and dialogue with their audiences, since, as has been shown, this is still very scarce.

It would also be interesting for the DMOs to commit to sharing on Twitter those reactions and positive comments generated by the public so that not only the mechanisms for dialogue and interaction would be reinforced, but also a type of promotion closer to the emotional values of the brand would be chosen.

In fact, although progress has been detected in most destinations, it is concluded that Twitter continues to be used more as an information channel than as a vehicle for interaction. This type of use is reflected in the number of followers: the national account of Spain and the regional accounts of Portugal, which are the ones that use this type of resource the most, are the ones with the fewest followers. We thus agree with Mariani, Mura, and Felice (2018), who had underlined that, until now, DMOs do not usually give much space to user-generated content.

On the other hand, the results show that the activity of tourist destinations on Twitter is not as closely linked to the number of followers on this social media platform as the size of the destination itself. In fact, the destinations with the most followers are not the ones that publish the most tweets or even those whose posts have a higher performance index.

The initial hypothesis that destinations maintain little interaction with their audiences has also been confirmed, although there are notable differences between destinations in Spain and Portugal. In fact, it is the Portuguese destinations that have a higher level of interaction (mainly Algarve, Azores, Lisbon, and Turismo do Centro).

On the other hand, since, through their tweets, users can play an important role in creating the brand image of destinations, DMOs should be more willing to dialogue with their audiences and actively listen to their needs. Thus, it would be advisable to increase the level of participation in the conversations that are generated on Twitter, thus preventing this social media platform from being used only to inform.

To this end, the DMOs should make a significant commitment to creativity and innovation so that they can use the information received from travelers to respond to the needs they may have, reinforce their experiences in the destination, and, therefore, contribute to the construction of a better tourist image, especially when it could be weakened, as has happened during the COVID-19 pandemic.

REFERENCES