How the COVID-19 pandemic changed forest visits in Switzerland: Is there a back to normal ?

collected two weeks after the lockdown was enforced (2020), and also with data from the summer of 2021. We aimed to determine if forest visitation had returned to normal by the time of the third survey. Our findings indicate that during times of crisis, visitation patterns and recreational preferences in urban green spaces, such as forests, undergo changes, with some of these changes exhibiting greater persistence than others. Meanwhile the activities during a forest visit as well as motivations of forest visits seem to return to their initial levels after the crisis, for example, the changes in the frequency or lengths of the visits appear to be more persistent. Recognizing and understanding these changes are crucial for urban planners and forest management authorities to adapt policies and effectively address the evolving needs of visitors.

When the World Health Organization (WHO) declared COVID-19 a pandemic on March 11 2020(World Health Organization, 2020), the importance of UGSs became even clearer.With restrictions on social life and diverse additional pandemic-driven stressors, nearby forests, parks and other green spaces were highly visited, for their positive effects on mental health and for outdoor recreation.For example, Venter, Barton, Gundersen, Figari, and Nowell (2021) used data from 53,000 STRAVA (an internet application for tracking physical exercise) users in Oslo and found that visits to UGSs increased in spring 2020, but it remained an open question as to whether the increase in recreation patterns persisted.Likewise, Ugolini, Massetti, Pearlmutter, and Sanesi (2021) reported that visiting habits to UGSs in Italian cities changed significantly during the first phase of the pandemic.During the lockdown, walking the dog, the need to relax, and physical exercise became increasingly importantalso depending on the infection rate in the cities.Additionally, the proximity of UGSs became more and more important (see Ugolini et al. (2020) as well).Considering longer-term patterns, Noszczyk, Gorzelany, Kukulska-Koziel, and Hernik (2022) analyzed the effects of the lockdown on UGS visits for the city of Kraków (Poland) from March to November 2020.They observed that the number of UGS visitors fell by 13.1 % compared with before the pandemic and that with each phase of the pandemic the percentage of people refraining from visits increased.Similarly, Jay et al. (2022) analyzed smartphone mobility data for 11,890 city parks in the U.S. and reported that visits declined by 36 % from March through November 2020.However, they found that there were rebounds when restrictions were relaxed.
For forests in particular the situation was similar.Ciesielski, Tkaczyk, Hycza, and Taczanowska (2022) distinguished between several subperiods of the pandemic year (2020) to investigate the effects on outdoor recreation in forests.They found that visitation patterns depended strongly on the sub-period and study area, but overall they concluded that aggregated data on visitor loads showed an increase in visitations in two regions of Poland.However, the most frequented recreational destination (located in the Kozienice Forest District) experienced a decline in visitation numbers in 2020 in comparison to 2019.Meanwhile, Derks, Giessen, and Winkel (2020) analyzed automated visitor counts for Bonn, a medium-sized city in Germany, and reported that visitor numbers in forests more than doubled in the period from 22 March to 30 April 2020.Likewise, Weinbrenner et al. (2021) found that people in Freiburganother medium-sized city in Germanyvisited forests more often during the lockdown in the first phase of the pandemic.Easier compliance with social distancing in forests, as well as the need for rest and retreat, were cited as the main reasons.In a study of forests in Switzerland, Wunderlich, Salak, Hegetschweiler, Bauer, and Hunziker (2023) observed that for people who visited the forests both before and during the lockdown, the frequency of visits increased during the early lockdown phase while the duration of visits decreased.The home-office work situation was one of the decisive factors for the increased frequency of forest visits by this group.
Altogether, studies on the effects of the lockdown on UGS visits in the first phase of the pandemic mostly indicate a higher demand for green spaces as their benefits for healthy outdoor recreation became more important (see also Twohig Although the pandemic clearly increased visits to some UGSs, the aforementioned studies have not been conducted to analyze the post-COVID situation.
We are in a unique position to analyze visitation rates to forests in Switzerland by the same cohort in the pre-pandemic phase in March 2020, during the lockdown phase in April 2020, and then again in the rather post-pandemic phase in June 2021.Even though, in June 2021 COVID-19 was still presentwith rising vaccination rates, we argue that the situation in the summer of 2021 was already rather "postpandemic".
By surveying the same respondents before, during and after the COVID-19 crisis, we can assess if the rising COVID-19 cases led to a higher frequency and longer duration of forest visits.Further, such an analysis eliminates the bias that an ex post facto evaluation may carry (Edmonds & Kennedy (2017), see also Wunderlich et al. (2023)).As the COVID-19 pandemic was prevalent worldwide and governmental responses were similar in most countries, the results of the study may contribute to a broader understanding of how permanent the social and spatial changes in visits to UGSs (in the following urban and peri-urban forests are considered) have been at an international scale.
Specifically, we aim to address the following research questions: 1. How did the frequency of visits to forest UGSs in Switzerland evolve during the pandemic?Was the initial level (from before the pandemic) reached again?Has there been a back to normal? 2. Did the spatial pattern of forest visitations change from before, during, and after COVID-19?Which sites were most visited before, during and after the pandemic?3. Was there a change in the activities during a forest visit?And in this respect: Was there a change in the rationale underlying a forest visit? 4. How did the relaxation status after a forest visit change from before, during and after the pandemic?Were there any changes in the perceived level of disturbance during a forest visit?
With our study we aim to enhance the understanding of the impact of the COVID-19 pandemic on forest management, particularly in relation to the significance of UGSs.By examining changes in forest visits from before, during and after the pandemic, we provide valuable insights for urban planning and policy development.This understanding can guide the design and management of UGSs ensuring equitable, livable and resilient cities.Our research addresses a literature gap and highlights the relevance of UGSs in satisfying human needs and integrating citizen behavior into future planning and management.Overall, our study has implications for both pandemic and post-pandemic urban planning and policy, contributing to the development of informed decisions on UGSs design and management.

Study region
Switzerland is a landlocked country in Western Europe.Municipalities are classified as urban, peri-urban, or rural based on population density.The majority of the population lives in urban (more than 10,000 inhabitants) or peri-urban areas (2,000 to 10,0000 inhabitants) (40 % of the area), with the largest urban areas being Zurich, Geneva, Basel and Lausanne.Only about a quarter of the population resides in rural municipalities (fewer than 2,000 inhabitants) (60 % of the area).The country is characterized by the presence of the Jura and Alps (including Prealps) mountain ranges, which include smaller towns and villages in rural and peri-urban areas (Fig. 11 in the Appendix).

Survey design and data management
In Switzerland, a nation-wide online survey (Swiss Socio-cultural Forest Monitoring WaMos) on forest visits and respondents' attitudes towards the forest is conducted every 10 years (BUWAL, 1999;Hunziker, von Lindern, Bauer, & Frick, 2012;BAFU & WSL, 2013;Hegetschweiler et al., 2022;BAFU, 2022).The most recent survey, with n = 8, 064 respondents, ended on March 9 th 2020 (t 0 ) -right before the COVID-19-induced lockdown and right before COVID-19 became ubiquitous.To study the impact of the lockdown and rising COVID-19 infection rates, we commissioned another survey from April 2 nd 2020 to April 9 th 2020, asking some of the same questions again (t 1 ).This time n = 1, 085 respondents from the original pool of n = 8, 064 respondents answered the questionnaire (see Wunderlich et al. (2023)).During this time restrictions in Switzerland were strict: Places of social encounters (e.g., shops, restaurants or leisure facilities) were closed.Social gatherings were restricted and the number of people allowed to attend private events was limited, beginning on 16 March 2020.The Federal Council also recommended to stay at home and work from home; shorttime working regulations were extended during this time as well with more than 1,000,000 short-time workers in April 2021 (Federal Office of Public Health, 2020a;Brülhart, Lalive, Lehmann, & Siegenthaler, 2020;State Secretariat for Economic Affairs, 2020).About 50 % of the Swiss employees were able to work from home during this time (Melian, Laude, & and Grampp, 2020).
In the context of decreasing COVID-19 infection rates on the one hand and increasing vaccination rates on the other hand, we then conducted the same survey once again from June 17 th to June 30 th 2021.Of the n = 1, 085 respondents from the second survey, n = 734 participants answered the questionnaire again (t 2 ).During this period, Switzerland found itself in a post-pandemic scenario as the mandatory home office requirement was lifted and replaced with a recommendation.Further, there were not guest limits in restaurants and large events were allowed again (with COVID certificates).Cultural and sports activities were mask and distancing-free.Discos were reopening and quarantine was exempted (Federal Office of Public Health, 2021).
Overall, all participants were members of the representative online panel of the Swiss market research institute LINK, which conducted the survey.The institute sent a link to the questionnaire to their panel members until given quotas regarding language, region, age and gender were filled.Quotas were based on census data of the Swiss population from the Swiss Federal Office of Statistics.Small incentives were given for participation.The questionnaire was available in three of the Swiss national languages (German, French and Italian) and could be filled in by using a computer or mobile device.LINK provided survey weights based on the base weights, accounting for probabilities of being selected into the sample, and on post-stratification weights, which were in line with benchmark distributions from Switzerland Census Population Surveys from 2019.
Table 1 in the results section shows the representativeness of our sample with a high proportion of respondents living in highly populated cantons such as Zurich.21 % of the respondents lived in a rural area; 79 % in an urbanized area or city (see Wunderlich et al. (2023)).
In the first wave of the survey, respondents were asked about their total number of forest visits during the spring, summer and autumn and instructed to categorize their visit frequency, ranging from "almost daily" to "never" (5 categories overall).Annual visit numbers were also reported.The second wave focused on visit frequency since the introduction of the lockdown, using the same categories.In the third wave, respondents reported their forest visit frequency after May 31 st , along with current visit numbers.This information allowed us to calculate weekly visit frequencies for all three periods.Respondents also provided data on their visit durations and most frequently visited forest locations using an interactive online map, where they circled the forest area they visited most frequently.Travel distance, motivations, activities and disturbances during visits were also recordedvariables that are likely to influence the frequency and duration of forest visits (Hunziker et al., 2012).The survey data was merged with COVID-19 prevalence data by canton Federal Office of Public Health (2020b) (for further information on the survey and on Swiss COVID-19 prevalence see Wunderlich et al. (2023)).

Data analysis
To analyze the spatial effects (PPGIS analysis) of the COVID-19 lockdown, we asked respondents of the survey to mark a point on an interactive map of the forest that had been visited most frequently.Hereby, we took into consideration all urban or peri-urban forests in Switzerland.Respondents were asked to place a circle in the designated forest area and were able to zoom in and out of the map and move it as needed.Further, respondents were able to switch between the map view and satellite view.
Then we merged individually mapped spatial points with survey responses using unique IDs.We focused on 448 respondents who visited the forest once or twice a month before, during and after the lockdown.First, we calculated the total number of data points in each 5 x 5 km raster cell (COUNT) out of a total of 3080 cells.Then, we summed the individual visitation frequencies per raster cell (SUM).We used the COUNT variable as a weighting factor for the visitation frequency (SUM) to account for the importance of raster cells with more observations (for each period separately).By subtracting the pre-COVID-19 situation from the lockdown and post-COVID situations, we estimated the changes in the spatial distribution of forest recreation.We also computed Pearson correlation coefficients of the differences between the pre-COVID and lockdown phases, between the pre-COVID and post-COVID phases, and between the lockdown and post-COVID phases to determine the significance of these changes.
In this study we utilized the panel data from all three periods (pre-, during and post-pandemic) to assess the relationship between visit frequency and independent variables.
To investigate whether visitor frequency increased during the public health emergency, we considered several independent variables: (i) COVID-19, a dummy variable indicating the emergence of the pandemic; (ii) ACTIVE, representing the number of active COVID-19 infections in Switzerland during the survey periods, including its squared term; and (iii) DEATHS, denoting the cumulative number of COVID-19 deaths reported until the beginning of each survey period, including its squared term.We include each squared term to account for potential non-linear relationships.This decision stems from the anticipation that the effect of active infections on visitor frequency may not follow a linear pattern.As the number of infections rises, the perceived threat might escalate, but at a certain point, habituation effects or advancements in vaccines and treatment options could lead to a diminishing impact or even a reversal of the trend.Including the squared terms allows us to capture potential curvature or diminishing returns in the relationships between these variables and visitor frequency, contributing to a more nuanced understanding of the dynamics during the public health emergency.Additional control variables encompassed socio-demographic factors (age, gender, educational background), characteristics of the residential area, work status (including homeoffice opportunities and short-time work), Swiss language area, behavioral variables (motivations and activities during a forest visit), travel time to forests, means of transportation to the forest, and the extent of feelings of disturbance during visits.We used random effects models and fixed effects models, presenting the fixed effects models based on Hausman test results.As the frequency variable is censored at 7 days per week, we used panel Tobit models.However, due to limitations of the plm package in R, we only conducted random effects estimations.

Results
Of the total number of responses analyzed (n = 734), 47 % were from male participants (53 % female).The mean respondent age was 55 years, while the average age for the Swiss population over 18 was around 51 in 2019 (Federal Statistical Office, 2020).Respondents resided across all 26 of Switzerland's cantons, with a large share of them located in highly populated cantons in Switzerland, such as Zurich.Further, 79 % of the respondents came from urbanized areas or cities (21 % from rural areas).Finally, the sample consisted of 7 % Italianspeaking respondents, while 22 % came from the French-speaking part of Switzerland and the remaining 71 % from the German-speaking part of the country (see also (Wunderlich et al., 2023)).

Frequency of forest visits
COVID-19 changed visitation patterns to the forest most frequently visited but also activities during a forest visit, motivations for a visit, and the relaxation level after a visit.During the first survey (before COVID-19), n = 615 of the n = 734 respondents visited the forest once or twice a month.During the lockdown, only n = 446 respondents visited the forest at least this often.In the post-COVID period, the number of frequent forest visitors increased again to n = 518.Fig. 1 shows forest visit frequencies to the most frequently vistied forest before (t 0 ), during (t 1 ) and after (t 2 ) the pandemic.The proportion of visitors who never visited the forests (1.5 %) increased during the pandemic (to 26 %) and did not fully recover in the post-COVID period.Most visitors who visited approximately once or twice a month before the pandemic stopped their visits during the lockdown phase with some of them visiting once or twice after the pandemic again.Overall, after the pandemic this visitor group is smaller in comparison to before the pandemic as well.Visitors who went to the forest once or twice a week before the pandemic increased their visits during the pandemic, but values returned almost to pre-pandemic levels during the post-COVID period.The proportion of visitors going once or twice a week was 37 % before the pandemic, decreased to 30 % during the pandemic, and then increased to 35 % in the post-COVID period.Among visitors who went less frequently than every second week (or: "no more than twice a month" in Fig. 1) the proportion decreased from 15 % before the pandemic to 13 % during lockdown, then increased to 20 % in the post-COVID period, surpassing the pre-pandemic level.The percentage of daily visitors during the pandemic was 19 %, which decreased to 16 % in the post-COVID period, nearly returning to the initial level of 15 %.However, pairwise t-tests did not show significant differences in visitation patterns between the different time periods (for t 0 and t 1 ,t = 0.054,p = 0.957; for t 0 and t 2 ,t = 1.038, p = 0.2996; for t 1 and t 2 , t = 0.97, p = 0.33).
Table 2 in the Appendix lists the results of the regression with forest visit days per week as the dependent variable for the fixed effect models; Table 3 in the Appendix lists the corresponding results of the tobit models.In each table, we first report the results from the model (1) including the COVID dummy.To explore a potential habituation effect among the forest visitors, we also present two specifications with the two alternative independent variables (2) ACTIVE and (3) DEATHS.We only show the statistically significant results.We observe a robust and statistically significant positive effect of COVID-19 on the number of forest visits per week, indicating that forest visits increased during the pandemic.The results further indicate that travel time played a significant role in forest visit frequency: longer travel times led to fewer visits.Some mandatory activitieslike dog walkingshowed a robust and statistically significant positive effect as well.The same applied to various motivations for visiting the forest in the tobit models (Table 3 in the Appendix)."Obligatory activities" such as dog walking and horse riding had a significant positive effect on forest visits per week as well.The opportunity to work from home also seemed to increase forest visit frequency, while being unemployed or in short-time work decreased forest visits.
Overall, the relaxation level after a visit seemed to be of importance as well, indicating an increased number of forest visits resulting in higher relaxation levels.This variable was significant for all models except for the fixed effects model using DEATH as a measure for COVID-19.This means that the relationship between the number of forest visits  and the relaxation level was driven by the number of active cases rather than by the number of deaths.Nevertheless, in the tobit models, the variable was significant in all three models.

Visit duration and travel time
The regression results show that travel time had a significant negative effect on the frequency of forest visits to the forests visited most frequently.importance during the pandemic: the proportion of respondents visiting forests with a travel time equal to or less than 5 min increased during the pandemic, from 29 % to 43 %, then declined to 37 % in post-COVID times.Likewise, shorter travel times of 6-10 min increased persistently, from around 29 % before the pandemic to 34 % during and 36 % after it.Meanwhile, longer travel times were avoided during the pandemica trend continuing in post-COVID times (Fig. 2).This indicates that the population continued to visit different forest patches in the post-COVID period, and that the changes appeared to be persistent.These differences between before and during the pandemic, as well as between before and after the pandemic, were highly significant (t 0 and t 1 , t = 4.62, p < 0.001; t 0 and t 2 , t = 4.16, p < 0.001), while changes between during and after the pandemic were not significant (t 1 and t 2 , t = − 0.956, p = 0.34).
In addition, we observed persistent changes in the duration of forest visits.Before the pandemic, only 19 % of respondents made shorter visits of less than 30 min.During the pandemic, 33 % preferred shorter visits, and in the post-COVID period 29 % still visited the forests for less than 30 min (Fig. 3).The differences between t 0 and t 1 were significant (t = 4.39, p < 0.001), while the differences between t 0 and t 2 (t = 1.33, p = 0.18) and between t 1 and t 2 (t = − 1.70, p = 0.089) were not significant.

The most popular forests
The GIS analysis underlines the trend of visiting nearby forests with shorter travel distances.However, overall, the visiting patterns realigned towards the pre-COVID period but at a slightly lower level.The spatial pattern of visitation frequencies across the time periods remained mostly unchanged on the national scale, according to a one-way ANOVA, which did not show statistically significant differences between the periods (F(2,771)=0.797,p = 0.451; Fig. 4).Forest recreation spots with high visitor frequencies were predominantly located in agglomeration areas, situated between the northwestern mid-elevation Jura ridges and the central to southern mid-to high-elevation Prealps and Alps.Highly frequented forests were in proximity to major cities (see highlight 1 on map A in Fig. 4) and in areas such as the region between Zurich and Aarau, (2) around St. Gallen, Solothurn (3) and Chur (4), and the southern part of Ticino at the border to Italy (Chiasso) (5).These areas mostly comprise urban and peri-urban community types (see Fig. 11).Differences in the weighted visitation frequencies across the time periods are presented in Fig. 5. Compared with the pre-COVID period, average frequencies tended to be lower in the post-COVID period.While highly visited recreation spots still remained popular, the weighted average visitation frequency was lower in the post-COVID period.However, this decrease in average frequency was likely balanced out by surrounding areas, emphasizing the importance of the mapped areas for everyday recreation.In certain areas, particularly (1) between Aarau and Zurich, in and around (2) St. Gallen and (3) Solothurn, and in the inner-alpine valley around Chur (4), the visitation frequency in the post-COVID period had not yet reached the pre-COVID level.A one-way ANOVA indicated significant differences among the time periods (F (2,596)=3.371,p = 0.035).However, a follow-up Tukey honestly significant difference (HSD) post hoc test revealed that the differences only occurred between the COVID lockdown and post-COVID periods.This suggests that forest visitation frequency temporarily changed during the COVID-19 lockdown in Switzerland but subsequently realigned towards values in the pre-COVID period, albeit at a slightly lower level.

Changes in activities and motivations
We observed changes in the activities pursued during a forest visit (Fig. 6).Due to the restrictions on social life during the lockdown period, activities such as socializing during a barbecue were drastically reduced.Before the pandemic 23 % of the respondents used their forest visits for social gatherings like barbecues, while only 2 % had a barbecue in the forest during the COVID-19 lockdown.The initial pre-pandemic level had not yet been reached by June 2021, with only 12 % of respondents having barbecues during forest visits.Similarly, playing with children in the forest decreased from 17 % of respondents before the pandemic to only 6 % during lockdown and 9 % after the pandemic.Activities such as enjoying peace and observing nature also decreased during the pandemic and did not reach the initial levels after the pandemic.Dog walking was the most robust activity and was almost unaffected by the pandemic, with 16 % to 17 % of respondents walking their dogs in forests.Biking activities in forestsregardless of the kind of bike usedincreased after the pandemic (e-bike +6 %, mountain bike +3 %, (city bike) bicycle +3 %).Overall, all differences were highly statistically significant.
In line with this result, the motivation "I want to be active in sports" became more important for the respondents during and after the pandemic.Likewise, the motivation "I would like to do something for my health" became more important.Nevertheless, the differences for those items were not significant (see 4 in the Appendix).Being alone became less important as a motivation to visit the forest.The changes from the pre-COVID-19 period to the post-COVID-19 period were significant for this item (see Fig. 7).

Changes in relaxation level and disturbances
The relaxation level achieved after a forest visit changed considerably during the pandemic.When asked before the pandemic, the majority of respondents reported that they would be slightly more (44 %) or   even much more (43 %) relaxed after visiting a forest.During the pandemic only 33 % were slightly more or much more (37 %) relaxed after a forest visit.The level of relaxation reached after a forest visit had not yet reached its initial level at the time of the post-COVID survey: 38 % and 33 % of respondents were slightly or much more relaxed after a forest visit, respectively.There was a large increase in the number of visitors who felt equally relaxed before and after a forest visit from before (8 %) to during and after (27 %) the pandemic.The differences pre-COVID and during COVID times were highly significant (t = 6.37, p < 0.001).Differences between pre-COVID and post-COVID times were highly significant as well (t = 7.43, p < 0.001), while differences between during and post-COVID times were not significant (t = − 0.83, p = 0.41) (see Fig. 8).
The perceived level of disturbance during forest visits decreased during the pandemic.Even after the pandemic, the disturbance level remained lower in comparison to before the pandemic.Before the pandemic around 52 % of the forest visitors declared not feeling disturbed during a forest visit.This number rose to 70 % during the pandemic and to 75 % after the pandemic (Fig. 9).A Chi-squared test of association indicated that all the tested effects were highly positively significant (t 0 and t 1 , t = 64.25,p < 0.001; t 0 and t 2 , t = 119.24,p < 0.001; t 1 and t 2 , t = 90.9,p < 0.001) (see Fig. 10).
During the pandemic one of the main factors contributing to feeling disturbed during a forest visit was the number of other visitors present in the forest.Before and after the pandemic around 27 % were disturbed by the number of other visitors; during the lockdown almost half of the visitors felt disturbed by other visitors.Further, other walkers and hikers disturbed more visitors during (24 %) and after the pandemic (13 %) compared with before the pandemic (2 %).In line with this finding, but probably also because fast respiration was discussed as a main driver of COVID infections, the number of visitors who felt disturbed by jogging visitors increased as well, from 6 % to 21 % during the pandemic (7 % after the pandemic).Table 5 in the Appendix shows the significance levels of the different disturbance factors during a forest visit for the different time pairs.
In addition to trash and vandalismthe main reasons for feeling disturbed during a forest visit before the pandemicpeople on bikes were a source of disruption.Although biking was one of the activities that increased during the pandemic, the feeling of being disturbed by bikers actually decreased significantly during the pandemic and did not reach initial levels by the time of the post-COVID-19 survey.

Discussion
During times of crisis, like during the COVID-19 pandemic, the   et al., 2020;Rice & Pan, 2020;Pichlerová et al., 2023;Jarský et al., 2022).However, before our study it was unclear whether the trend of rising forest visits continued after the pandemic or if there was a return to "normal".In Switzerland, the distribution of visit frequencies underwent significant changes during different periods.During the lockdown, the proportion of visitors who never visited forests increased to one-quarter, not yet reaching pre-pandemic levels.Conversely, the proportion of respondents who visited the forest on a daily basis increased during the lockdown but declined in the post-pandemic period, approaching pre-pandemic levels.

Proximity to forests
The travel time to the forest that is most frequently visited decreased during the pandemic, with longer travel times negatively impacting visit frequencies (see 2 and 3).The proportion of visitors with a travel time of 5 min or less increased during the lockdown, indicating that visitors sought out different (closer) forests during the crisis.The PPGIS analysis supported these findings, highlighting areas such as those between Aarau and Zurich, around St. Gallen and Solothurn, and the inner-alpine valley around Chur, where visitation frequencies in the post-COVID period had not yet returned to pre-pandemic levels.In comparison, Noszczyk et al. (2022) found that the most popular green spaces in Krakow (Poland) were located in the heavily urbanized center of the city, where the green belt of the "Planty Gardens" encircles the main square.Further, they reported that parks situated along the Vistula river, which flows through Krakow, were popular.As in our study, the proximity of people's homes to green spaces influenced the frequency of visits during the lockdown, but the authors argued that historical value, cultural heritage, and the promotion of a sense of identity all played roles, too.Similarly, Ugolini et al. (2021) found that accessibility and the proximity of forests played a large role in visiting behavior.However, in that study the desire to visit a less crowded forest drove the forest visiting behavior as well.The chance to engage in remote work appeared to elevate the frequency of visits to the forest, whereas being unemployed or on short-time work was associated with a decline in forest visits.This is in line with the findings of Füzéki, Schröder, Groneberg, and Banzer (2021) that there was a smaller decline in leisure-timerelated physical activity among people working from home during the lockdown than among those not working from home.The result that working from home was a significant positive driver of forest visit frequency emphasizes the importance of proximity and spontaneity when visiting forests, as well as the significant role of travel time.Overall, we did not find support for the argument of Ugolini et al. (2021) that visitors started visiting more distant forests to engage in more physical exercise.On the contrary, in our study the overall duration of forest visits declined during the pandemic, with more visits being 30 min or shorter, a development that persisted in post-COVID times.In this respect, it seems that COVID-19 accelerated a trend that had already been observed in Switzerland since 2010 (Hunziker et al., 2012;Hegetschweiler et al., 2022).Proximity is discussed to be a barrier Signif.codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1. Signif.codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1.
Overall, the changes in forest visitation patterns during the COVID-19 crisis suggest complex potential long-term effects on forest utilization.

Activities and motivations
By altering the overall way of living, COVID had an impact on the motivations for and activities during forest visits, everything has not yet returned back to normal after the crisis.Due to policy restrictions on social gatherings, social activities like having a barbecue in the forest reduced enormously during COVID.In contrast to, for example, Grima et al. (2020), we did not find an increase in sport activitiesexcept for cyclingduring the pandemic.The most constant activity during the pandemic was dog walkingan essential activitywhile observing nature and enjoying peace became less relevant (similar to Ugolini et al. (2020Ugolini et al. ( , 2021))).Meanwhile, forest fruit collection declined in Switzerland, Jarský et al. (2022) find a significant increase in such activities during the pandemic for the Czech Republic.Nevertheless, the regression analysis shows that visitors who still pursued these activities (observing nature and enjoying peace) visited the forests more often during and after the pandemic.
Motivations related to the topic of health like, e.g., "doing something for my health" or "doing sports" became more important during the pandemic.Nevertheless, this change was not reflected in the activities during a forest visit (walking, bicycling, Nordic walking or running were activities that were less pursued during the pandemic) but supports the results of Grima et al. (2020).Overall, in post-pandemic times we observed a trend towards a back to normal with most activities and motivations, but the initial levels had not yet been reached.The exceptions were biking and e-biking, where post-pandemic levels were higher than those in pre-COVID times.In the post-pandemic period, the motivations of sports and health still seem to play a larger role in forest visits than in pre-pandemic times.

Relaxation and disturbances
A vast amount of research shows the positive effects of interaction and contact with nature on human health and well-being (Navarrete-Hernandez & Laffan, 2019; Kim & Miller, 2019;Schwartz, Dodds, O'Neil-Dunne, Danforth, & Ricketts, 2019).Before the pandemic, only about one out of ten respondents did not feel at least slightly more relaxed after having visited a forest.After-visit relaxation levels were not as high anymore during the pandemic, and initial levels had not been reached in the post-COVID-19 period.This finding is not in line with those from other studies; for instance, Noszczyk et al. (2022) reported that visitors believed their visits to UGSs had a very large or large influence on -their stress levels.Further, Pichlerová et al. (2023) found an increase with stated stress reduction after a forest visit during the pandemic although there was no significant effect of improved subjective well-being from number of forest visits.Lamatungga et al. (2024) showed that walking in a forest setting during COVID-19 significantly improved the cognitive function and increased heart rate variability in older adults in comparison to walking in an urban setting.
For Switzerland, it could be argued that shorter visits were the main reason that pre-pandemic levels of relaxation had not yet been reached, i.e. the forest visits simply were not long enough.This trend seems to be persistent in post-COVID times, but it started even before the pandemic in 2010 (see Hunziker et al. (2012); Hegetschweiler et al. (2022)).
One of our main findings is that the feeling of being disturbed during a forest visit decreased during the lockdown.Overall, this could be a result of the "we can do it" mentality during the lockdown and the feeling of being a part of something larger, in following social distancing orders.This result is similar to that of Weinbrenner et al. (2021), who found that forest visitors in Freiburg (Germany) did not perceive more disturbances than usual during lockdown, but it is in contrast to findings from other studies, e.g.Derks et al. (2020);McGinlay et al. (2020) reported increasing conflicts between recreationists during COVID.They argued that the fear of virus transmission and crowding of forests could have caused additional conflicts.In our study, however, the feeling of being less disturbed persisted even in the post-COVID period.Therefore, we argue that our result is consistent with, for example, Rugel, Carpiano, Henderson, and Brauer (2019), who found that being outdoors, including spending time outdoors near home, can be beneficial to the community because it contributes to feelings of solidarity.Beyond the feeling of being part of something larger during the lockdown, this feeling seems to be even stronger now.This in turn could have positive effects on stress-related illnesses and improve overall health and wellbeing (Bertram & Rehdanz, 2015;Carrus et al., 2015).

Implications for urban planning
Overall, our study highlights the importance of keeping track of visitor patterns in forestsespecially on the part of forest managers and urban planners.Visitor patterns to UGSs are not constant but rather subject to constant changes.So far, it seems that COVID altered forest visitation patterns not only during the pandemic but in a sustainable way.Therefore, adaptations in forest management are essential not only during a pandemic itself but also in post-pandemic times, and it is important for forest owners and managers to be aware of these changes.Our findings are additionally important for policymakers and city planners.Policies that aim to match sustainability and resilience goals, indicated by the 11 h Global Sustainable Goals (UN, 2015), should include plans for buildings and green spaces that foster human wellbeing, a healthy environment, social equality, and personal security.Proximity to UGSs should therefore be one of their main goals as well.
Our study underscores the significance of forests as versatile spaces for a range of recreational activities, encompassing social gatherings, relaxation, and solitude.This became particularly pronounced during the pandemic, wherein individuals sought outdoor pursuits while adhering to social distancing guidelines.The observed shifts in forest visitor behavior are likely to hold implications for public infrastructure providers and recreation facilities in the long term.Lockdown measures, including quarantine and social distancing, are poised to play pivotal roles in future preparedness plans.A critical task in the times ahead will be to facilitate social distancing while mitigating potential adverse effects on both physical and mental health.Forests and other urban green spaces are poised to emerge as vital reservoirs of resilience and solace in this future context.In this broader context, urban nature assumes the role of a sanctuary and a wellspring of resilienceoffering spaces for well-being, physical exercise, and alternative avenues for socializing (D'Alessandro et al., 2020;Kleinschroth & Kowarik, 2020;Lai, Webster, Kumari, & and Sarkar, 2020;McCunn, 2021).It is imperative to explore alternative pandemic mitigation measures that facilitate social distancing while preserving the positive impacts on physical and mental health (Slater, Christiana, & Gustat, 2020).
The realignment of overall visitation patterns to pre-COVID norms is a noteworthy observation; however, it is essential to recognize a slight decrease in forest visit frequencies during the post-pandemic period compared to the pre-pandemic era.This suggests that the heightened interest in forest visits during the pandemic may not be entirely sustained in the long term.In response, urban planners play a crucial role in promoting sustainable visitation patterns by ensuring diverse recreational opportunities beyond the pandemic.Simultaneously, managing visitor capacities becomes imperative to prevent overcrowding and preserve the ecological integrity of forest areas.This is important especially because the pandemic showed an increased travel behavior to urban forests (see also Randler, Tryjanowski, Jokimäki, Kaisanlahti-Jokimäki, & Staller (2020)).Policymakers, collaborating with urban planners, should prioritize increasing the quantity of UGS and advocate for at least one comparatively large UGS per neighborhood.The provision of accessible green spaces within cities or in proximity to urban areas should be a main point for policy makers to meet the increasing demand for nature-based activities.
Managing visitor behavior will also be important in this respect, to handle the changes in visitation patterns.Overall, our study reveals a decrease in social activities during the pandemic, such as barbecues and playing with children.Conversely, biking activities increased during and after the pandemic.Urban planners can take these changes into account by providing infrastructure and facilities that support biking and other popular activities in forest areas, such as designated biking trails or picnic areas.Further, integrating amenities that cater to diverse user groups, including families and individuals of different age groups, could enhance the overall experience and promote inclusive outdoor recreation.While the issue of financing these expenses is still being discussed, it is crucial to uphold the Right of Public Access to ensure that all residents have equal opportunities to access UGSs.It is vital to prioritize conflict prevention and involve all relevant stakeholders in identifying effective solutions.

Limitations and future research
Our study is not without limitations.First, the results could have been affected by the warm weather during the time of the surveys.During the survey that took place during the lockdown phase, temperatures were exceptionally high for April.In addition, the initial goal of the first survey (before the pandemic) was not to investigate the impact of the COVID-19 lockdown on forest visits but to explore the relationship of the Swiss population with their forests.Despite this, we were able to use some parts of the survey again to analyze the recreational effects of the lockdown within the same cohort, as the lockdown was enforced a few weeks after the first wave of the survey (see also Wunderlich et al. (2023) which concentrates on the effects of the possibility of working from home and short-time working regulations on forest visits).Further, we used the same survey again to estimate if the lockdown effects were persistent.Our survey in June 2021 was conducted during a time when there was a sense of transitioning to a post-pandemic phase.Thus, the three surveys were conducted at three different points in time regarding cultural and political sentiments.Accordingly, the varying observations we were able to identify are (in part) significant which is not surprising.However, it is important to note that the pandemic was not fully resolved at that time.As a result, the visitation patterns and activity changes we observed may have undergone further modifications after that point.Thus, conducting another study would be beneficial, particularly to determine if activity changes, such as increased biking, have persisted.
Additionally, in future research it will be crucial to examine the interplay between age, gender, income, education, family structure, living conditions, and ethnic backgrounds.Understanding these factors is essential because if increased outdoor recreation were to primarily benefit those with higher education and income, it would indicate a socially unsustainable change and potentially exacerbate existing inequalities.While our primary focus was on comprehending the overall influence of pandemic-related restrictions on forest visits, rather than delving into the specifics of individual regulations during the lockdown phase, such as social restrictions or closed shops, there is a notable interest in analyzing these effects.This exploration could offer a more nuanced understanding of how distinct regulations, for example, impacted the younger generations' inclination to visit forests more or less.
Further, our analysis is focused on understanding the changes between the three time periods, specifically examining the impact of the lockdown phase and whether changes in forest visits are persistent.Additionally, we were interested in exploring the alterations with help of the PPGIS-data.As a result, we do not delve into correlations between the responses in this context.Further, the study does not delve into the underlying factors that support or oppse the frequency of these visitations.However, we acknowledge that such an analysis could be of interest and would be intriguing for future research.Regression models could be employed, incorporating a variety of physical and social variables that could potentially explain these patterns.

Conclusions
The COVID-19 pandemic and its policy implications altered our lives from one day to the next.The lockdown phase was characterized by restrictions in everyone's private lives and therefore also affected recreational activities.We studied the impact of the COVID-19 pandemic on forest visits in comparison to visits before the rise of the pandemic.
Among the conclusions that can be drawn is that visitation patterns and recreational needs in UGSs change during times of crisis, and that some changes appear to be more persistent than others.For example, activities and motivations seem to return to their initial levels after the crisis has subsided.Therefore, it is possible that in the long run the initial levels will be reached again.However, it is likely that the increased visitation by bikers will be persistent.This, in turn, could lead to more conflict situations between different types of recreational users.Further, we conclude that the accessibility of forests plays an important role in the frequency of visits, evenor especiallyin times of crisis.
Even though the level of disturbance perceived by forest visitors was still lower at the time of the third survey in comparison to pre-COVID times, it is important to prevent new probable conflicts between forest users, as well as between forest users and forest functions, from the beginning.
Funding: The research for this paper was financially supported by the Swiss Federal Office for the Environment (FOEN) as part of the Swiss socio-cultural forest monitoring and as a contribution to monitoring forest recreation within the Swiss National Forest Inventory (NFI).

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
A.C.Wunderlich et al.

Fig. 4 .
Fig. 4. Weighted visitation frequency across Switzerland.Note: Map A represents the time point after the COVID-19 lockdown (June 2021), map B represents the status pre-lockdown (March 2020), and map C refers to a state within lockdown (April 2020).The highlighted red circles mark areas with an accumulation of high visitation frequencies.

Table 1
Representativeness of our sample.
A.C.Wunderlich et al.
Table 3 also indicates that nearby forests became of greater

Table 2
Regressions modeling the main effects on the change in forest visits: Fixed effect models.

Table 3
Regressions modeling the main effects on the change in forest visits: Tobit models.

Table 4
Significance levels of different motives of going to the forest for the different time pairs, using student's t-test.

Table 5
Significance levels of different disturbance factors during a forest visit for the different time pairs, using Chi2-test