Visitor frequencies and attitudes towards urban forests and their management, before and during the COVID-19 lockdown. A mixed methods case study in Bonn, Germany

ABSTRACT Urban forests play a crucial role for the wellbeing of city dwellers, and their importance for people has been emphasised during the COVID-19 pandemic. This exploratory study analyses the visit patterns and visitor attitudes and perceptions in a peri-urban forest nearby Bonn, Germany, as well as the impact of the lockdown. Methodically, we combined automated visitor counting with a total of 345 on-site interviews. Respondents were asked a variety of open-ended and closed questions on various aspects of forest management and recreation. The results show that shortly after the inception of the lockdown the number of forest visitors doubled and the visit pattern changed markedly. In contrast, people’s associations with the forest remained rather stable. The forest visitors interviewed primarily associated the forest with tranquillity, recreation and fresh air, and they were generally positive about forest management. However, these expectations conflicted with the sense of crowdedness experienced during the lockdown, when novel forest uses and new motivations for visiting the forest arose, with an important focus on the forest as a place for social interaction. These were mainly a result of the lockdown restrictions, rather than COVID-19 itself, which left people with more time and flexibility, and less alternative activities. The results highlight the importance of forest management in catering to people’s expectations and ultimately for the role that forests play for people’s wellbeing. This was the case before the lockdown but arguably even more so during, in response to a variety of needs resulting from unprecedented circumstances.


Introduction
The importance of forests for human wellbeing is widely acknowledged.They offer a multitude of ecosystem services for society, including cultural ecosystem services such as recreation, heritage value, spiritual enrichment and aesthetic pleasure (Bieling and Plieninger 2013;Torralba et al. 2020).
A recent review study by Ranacher et al. (2017) assessing how Europe's forests and their management are perceived by the general population, identifies forests primarily as places where people expect to find naturalness, beauty and relaxation, which can all be classified as cultural ecosystem services.Other research indicates that people like to visit the forest to get away from the hustle of everyday life, to experience what they perceive as 'wilderness' (cf.Rametsteiner and Kraxner 2003;Bauer and Körner 2005;Hunziker et al. 2013).Measuring cultural forest ecosystem services remains, however, difficult compared to provisioning, regulating and supporting forest ecosystem services.Yet, as Bieling and Plieninger (2013) point out, it may be cultural ecosystem services that are the most irreplaceable: wood can be imported, water filtration can be mechanised, but a scenic landscape cannot be replaced.
For Germany, a large-scale perception survey on forests indicates forest recreation as the main priority among the population (Wippermann and Wippermann 2010), ahead of environmental and economic concerns.So-called nature awareness studies, which are a recurring study in Germany, have highlighted the growing importance of forests as a place of wilderness, and the most recent edition shows that 94% of the respondents indicated feeling happy when being in nature (Nürnberg et al. 2019).Spending time in nature has been shown to lead to an enhanced sense of peace and place, and the feeling of 'wilderness' is thus not just a spatial construct but also a state of mind connected to that place (Kaplan and Kaplan 1989).
People not only feel happier in a green environment; they feel healthier as well (Pietilä et al. 2015).The link between nature and wellbeing has not only attracted interest from both the mental health and the physical health sector.There is a strong link between mental and physical wellbeing and trees play a vital role here, particularly through stress relief (Meyer and Bürger-Arndt 2014).Forest bathing for instance has been shown to reduce heart rate variability and blood pressure (Park et al. 2009;Roviello et al. 2021), and the (mostly) positive impacts of forests for mental health have repeatedly been shown (Bielinis et al. 2019;Kotera et al. 2022).Research further indicates that wellbeing is also connected to forest characteristics -e.g. the perceived level of biodiversity shows a positive correlation with reported wellbeing (Carrus et al. 2015).Clark and Stankey (1979) define the recreation opportunity setting of a forest as 'the combination of physical, biological, social, and managerial conditions that give value to a place'.Forest management -or the lack thereof -thus plays a crucial, but sometimes unnoticed, role in shaping recreation opportunities as it impacts the structural and species diversity of the landscape (cf.Edwards et al. 2012;Konczal et al. 2022).This is particularly important in the Central European context, where most forests are shaped by management, and the perception of forest management measures is thus a critical factor for recreation (Maier and Winke 2017).Yet, positive impacts on human well-being and health are arguably not the only motive behind the decision to visit a certain forest: as Meyer et al. (2019) find, habit, spontaneity and vicinity may even be more decisive.
In 2020, the capacity of forests to increase human wellbeing was put to the test.In March 2020, the World Health Organization (WHO) officially declared the COVID-19 outbreak as a global pandemic.Like many countries, Germany imposed measures to curb the spread of the virus, mainly physical distancing, restricting social contacts and closing non-essential shops, offices, sport facilities, sociocultural facilities and schools (Han et al. 2020).In the following section, this situation will be referred to as 'lockdown'.
The impact of these measures on human wellbeing has been subject to heated debates.Recent studies show some measures have had a negative impact on people's mental wellbeing, especially in vulnerable groups and right after the inception of the lockdown measures (Schwinger et al. 2020).There are indications that the lockdown has influenced forest recreation both in terms of visitor numbers (Derks et al. 2020;Venter et al. 2020) and in terms of attitude towards the role and the importance of forests and forest management (Hunziker 2020;Rice et al. 2020;da Schio et al. 2021).The emerging literature shows that in countries where forest visits remained possible, visitor numbers have greatly increased (Ugolini et al. 2020;Venter et al. 2021).Forest and other green spaces seem to have played a crucial role for people's mental health and wellbeing (Levinger et al. 2022;Stock et al. 2022).Jones et al. (2021) studied the impact of the increase in forest visits in 14 popular European National and Nature Parks and identified crowding and increasing conflicts as crucial challenges.A German case study found that the function visitors attributed to the forest stayed largely the same during the pandemic, but that the forest was increasingly spatially appropriated as a social space (Weinbrenner et al. 2021).People started seeing the forest as an extension of their living room or garden; by using it, it became 'theirs'.
At the start of our research however, nothing had yet been published on the effects of the lockdown on forest recreation.Our study originally set out to analyse the importance of the previously mentioned social and cultural functions in the Kottenforst, a large peri-urban forest in the densely populated region of Bonn, Germany.Visitor counting as well as a large set of interviews were conducted in 2019, well before the COVID-19 pandemic.When the lockdown was established in March 2020, a drastic increase in visitor numbers was observed in the forest (Derks et al. 2020).Hence, we decided to investigate how this increase was related to the attitudes of the visitors and spontaneously conducted a second round of interviews with forest visitors during the lockdown.
In this paper, we present both extended quantitative data generated by the visitor counting device covering a timeline from April 2019 until May 2022 and data on the visitors' views of and emotional bond with the forest gathered with two surveys; one before and one during the first strict lockdown, which this paper considers the period between mid-March 2020 and early May 2020.
Our research questions are:

Case study description
Our study was initiated in the Kottenforst, a periurban forest area near Bonn, in 2019 in the framework of a LIFE+ project (LIFE13 NAT/DE/000147).
The focus of that project was to study the effect of the restoration of degraded forest habitats on visitors' perceptions towards the forest.The Kottenforst stretches over more than 5000 ha, immediately bordering the former German capital city Bonn, and is located south of Cologne in North-Rhine Westphalia.With 815 inhabitants per km 2 , the Bonn-Cologne region is rather densely populated.The Kottenforst lies at the transition between the industrialised and urbanised lower Rhine Valley and the less populated undulating hinterland to the southwest.It is mostly owned by the state of North-Rhine Westphalia, but the northern part belongs to the city of Bonn and some smaller areas are private.
The forest is situated on a plateau overlooking the city and consists of a variety of tree species, most notably oak (Quercus robur and Q. petraea) at 38%, beech (Fagus sylvatica) at 19% and Norway spruce (Picea abies) at 15%.Much of the forest is part of the European Natura 2000 network, classified as oakhornbeam forest (Habitat Type 9160, Stellario-Carpinetum) and beech forest (Habitat Type 9110, Luzulo-Fagenion).Even though the Kottenforst is mainly characterised by native tree species (except for some spruce plantations mostly planted in the first half of the 20 th century), forest structure and composition has been strongly shaped by a long history of human use and management.The Kottenforst is a prime example of a large peri-urban forest in Europe that has been and is used in multiple ways, and provides multiple ecosystem services.Recreation and biodiversity conservation are of high importance.In management, these objectives are combined with sustainable wood harvest in an approach that has been labelled 'integrated forest management' (Kraus and Krumm 2013;Aggestam et al. 2020).Specifically, the forest management is characterised by continuous cover forestry with selective cuts of valuable stems.However, bark beetle infestations in recent years have led to clear-cutting some (dying) spruce stands, which has a large visual impact and has been much discussed in the local press (Schölmerich and Striepen 2020).In turn, only a small portion of the forest is strictly protected: 5% is exempted from active management (Schölmerich and Striepen 2020).
When the COVID-19 pandemic hit Germany, the federal state ordered the limitation of social contacts starting 22 March 2020.This period, which ended in May 2020, will be referred to as the 'first lockdown' in this paper.Bars, restaurants, schools, sport clubs and non-essential businesses were closed.Employees were encouraged to work from home if possible, and people were only allowed to go outside with a maximum of one other person belonging to the same household.In early May, the most severe restrictions were lifted.Shops, schools and kindergartens were gradually reopened but social contacts and cultural events remained restricted until and even partially through the summer.A second lockdown in the autumn and winter of 2020/2021 was followed by the lifting of many measures again in spring and summer, followed by renewed (albeit softer) restrictions in the winter 2021/2022 and eventually a lifting of most measures in spring 2022.

Data collection
The study relies on a mixed method survey approach with a convergent parallel design.Specifically, it combines an automated visitor counting to assess the pattern of visitor numbers in the Kottenforst with 345 semi-structured interviews with forest visitors.Through this exploratory sequential approach, it assesses visitor frequencies over a longer period before and during the pandemic and aims to link it to an assessment of people's forest-related perceptions and motivations to visit the forest.

Visitor counting
With the aim of counting forest visitors, an infrared sensor was activated on 15 April 2019.The counts in this paper run up until the end of May 2022, when the device was relocated to another forest in the region.The device was installed at a carefully selected location on a gravel road next to a clearing in the forest, which was attractive to recreationists but also used by commuters.The sensor location represents a hybrid form of access corridor, somewhere in character between the (common) straight asphalt roads and the (less prevalent) winding dirt roads that are both present in the Kottenforst.The aim of this census was not to register the total amount of people that visit the Kottenforst during a given period, a task that would require more than one point of measurement, but rather to track the variations in intensity of forest use throughout the day and throughout the year.

Surveying visitor perceptions and motivations
In order to assess perceptions and motivations of forest visitors and how they related to various forest management interventions, 292 interviews were carried out between May 2019 and December 2019 and 53 between April 2020 and June 2020.The lower number of interviews during the second round was due to a number of reasons.The first set of interviews was planned in advance and carried out throughout the year.The second set of interviews was conducted ad hoc when noticing the surge in visit counts during an on-line analysis of our digital counting.They were undertaken during a short time frame, with no additional budget and under difficult circumstances, including restrictions on travel and meeting in groups.
The interviews were based on interview guidelines combining open and closed questions to record the visitors' perceptions of the cultural, recreational, and aesthetic ecosystem services provided by the forest, and to assess their attitudes towards forest management measures.Open questions related to (1) the reason for their forest visit, (2) main associations with the forest and (3) issues that disturb the visitors.Closed questions on a 5-point scale referred to (4) the perceived beauty and naturalness of the specific stand where the given interview took place (accompanied by an open question on which indicators this perception is based), ( 5) the opinion on felling trees for different predefined purposes and ( 6) the frequency of forest visits and -only for the interviews carried out during the lockdown-whether this has changed during the pandemic (combined with an open question on reasons for the changed frequency).
The interviewees were selected randomly as they passed by the interviewers.Interviews were conducted at many different times throughout the day and the week, from morning to evening and from Monday to Sunday.Since people were approached directly, the response rate was very high.Despite the lack of absolute numbers, we estimate that over three quarters of the people who were invited to participate were willing to spare some time for a short interview.Fifteen interview locations were selected together with the local forest managers, representing the diversity of stand types in the forest in terms of age, structure and species composition (see Figure 1).Some question pertained to the specific forest stand where the interview took place.A descriptive overview of the selected stands is provided as Appendix.
Finally, three forestry professionals who work in the Kottenforst were interviewed during the first lockdown to help interpret the data from the perspective of the local forest managers.

Data analysis
The visitor counts were analysed using simple descriptive statistics through a univariate analysis: visitor distribution was plotted against different time units in order to compare trends before and during the lockdown.
The results of the survey, including the qualitative answers to the open interview questions, were coded using an inductive coding strategy (cf.Edmonds and Kennedy ; Nie 2017).Each interview question was coded separately and as a result the answers were grouped into categories.The perception analysis comprised the complete set of interviews, but also differentiated between the before and during lockdown interviews.The findings relate directly to the interview data and provide a basis for potential future studies.
The quantitative part of the interviews was processed with descriptive statistics for most topics (acceptance of forest management vs. age, gender, and visit frequency), and by calculating correlation coefficients between given variables (perceived beauty and naturalness).The limited number of interviewees and the exploratory setup of the study limits the possibilities for statistical testing of the interview answers.Some statistical test were carried out, specifically the Spearman's correlation coefficient when comparing Likert scale results and t-testing for comparing coded answers between the groups.More importantly however, we tried to identify tendencies based on the face to face, on site interviews and observations that were made by the forestry professionals.Numbers and percentages expressed in the following sections should be interpreted in that light as illustrating trends, not as representative statistical findings.

Evolution of forest visits
The visitor counts allow for an analysis of the impact of the lockdowns on the number of forest visitors in the Kottenforst (Figure 2).Compared to the same period in 2019, the average number of daily visitors more than doubled during the first lockdown, from 298 to 627.The increase was less pronounced during the second lockdown (between 13 December 2020-10 March 2021) with an average of 322 daily counts compared to 219 the year before.The 'lockdown-free' summer of 2020 was the only period since the pandemic when visit counts were actually lower than in the pre-COVID year 2019.Since then, the numbers have stabilised but are still slightly above the pre-COVID observations.
During the lockdowns, the visitor distribution differed markedly from the pre-lockdown situation (Figure 3).In the year before the pandemic, visitor numbers during the week peaked just before and after office hours, showing a strong presence of people commuting to and from work, or conducting outdoor workouts.During the weekends, the curve looked different; the visitors were more spread out during the day, culminating in the early afternoon.In terms of average daily totals, Sunday was by far the most popular day to go to the forest, with Saturday being the least.During both lockdowns, the distribution patterns showed a curve reminiscent of the weekends before, but with a more marked increase in the afternoon.The difference between the weekends and weekdays became less pronounced.Sunday remained the most popular day, but the previously unpopular Saturday became almost equally busy.In general, people came to the forest later in the mornings and left earlier in the evenings.While the presence of commuters was still detectable during the first lockdown, it was almost completely absent during the second (Figure 3).
To be complete, a number of meteorological factors were also analysed.In general, the visit frequency was higher on warm or dry than on cold or rainy days.The clearest positive correlation was found between visitor numbers and solar irradiation.Weather patterns do not explain the significant longterm variations in visitor patterns throughout the pandemic, but they can contribute to explain variances from week to week.
The visitors were mostly middle-aged or elderly people (Table 1).The share of people under the age of 40 represented less than a fifth of the census group.Most visitors reside in one of the municipalities immediately adjacent to the forest.Men were slightly overrepresented compared to the municipal average.
Asked about the self-declared changes in visit frequency during the lockdown (N = 15), about one third of the interviewees of the second survey said they went to the forest more often during the lockdown than before.The majority, however, claimed not having changed their visit frequency.Furthermore, nearly 90% of all interviewees expected to visit the forest with the same frequency following the pandemic as before it, with only 10% responding that they expected to increase the frequency of forest visits in the longer term (Figure 4).
The reasons provided for the increase in forest visits by those who went more frequently during the lockdown (N = 15) were often rather pragmatic and largely related to the restrictive measures affecting respondents' leisure time (Figure 5).Some of the main recorded reasons were that people had more time and more flexibility due to their remote working situation.Health and contemplation was the second motive, followed by exercise and sports.
More respondents were accompanied by family members during the lockdown than before.Contact with people became a main reason to go outdoors more often, indicating a different appropriation of the forest as a living space and a place for social interaction.

Associations with the Kottenforst
Respondents were asked (open question) for three words they associated with the Kottenforst.Figure 5 and Table 2 give an overview of the results.
The recorded associations covered all the senses except taste: smells, sounds and touch were all mentioned.Aesthetic aspects were prevalent, as were associations linked to emotions.Key words that were coded under these two categories are shown in Table 2. Aspects related to nature were frequently mentioned by just over half of the sample group, but mostly on a general level.'Nature', 'forest' and 'trees' were the most common words.Only very few respondents (6%) mentioned specific tree and animal species, namely: 'oak' and 'beech' or 'bats', 'wild boar', 'deer' and 'tawny owl'.The fundamental associations that people had with the forest did not change much after the inception of the first lockdown: finding peace and quiet, fresh air and nature were still by far the main appealing factors of the Kottenforst (Figure 5).Note that tranquillity and fresh air only comprise one recorded term each.The other coded categories are compound codes consisting of several related utterances.The most notable changes were that during the lockdown less people mentioned fresh air and more associated the Kottenforst to nature.Also tranquillity was less emphasised during the lockdown, possibly related to the crowding of the forest as a result of doubled visitor numbers.Emotional aspects in general also seemed to have become more important.

The impact of forest management on recreation
At every interview point, the respondents were asked to express their appreciation for a specific forest stand at that point, by scoring the perceived naturalness and beauty of the stand on a 5-point Likert scale.In general, most stands scored highly on perceived beauty and only slightly lower on perceived naturalness.Forest stands  that were perceived as being more natural were generally also seen as the most visually attractive (Table 3).Specifically, a moderate but significant positive Spearman correlation coefficient of 0.436 was found between the respective scores for naturalness and beauty.
Interviewees were also asked to articulate the reasoning behind their judgement.Associations with naturalness and beauty largely aligned.In response to the open question about what three keywords describe the beauty of the stand, the main indicators were broadleaf trees and species mixture, diverse stand structure, wild look, tree health and green (Table 4).Emotional aspects, such as 'secretive atmosphere', 'romanticism' or 'the sound of the wind', were also found.
The perceived naturalness of the stand was mainly characterised by the absence of visible forest management, the presence of broadleaf trees and species mixture (Table 5).Diversity in terms of species and structure were thus characteristics that clearly linked perceived beauty and naturalness.
The colour 'green' as a predictor for beauty gained popularity during the lockdown.Also tree health became more prevalent.Similarly, 'green' also gained importance as a predictor for the perceived naturalness of a stand.Diversity in terms of species and structure was also seen as more crucial during the lockdown than before.
Interviewees were also asked about their opinion on felling trees.Despite the negative impact of visible human interventions on perceived naturalness, the proclaimed acceptance of felling was relatively high, with 63% of the interviewees agreeing with the overall idea.Sick spruce monocultures in the Kottenforst, widely infested with bark beetle, were viewed as the least attractive and least natural stand type.Accordingly, the acceptance of tree felling of such damaged or dying stands for phytosanitary reasons was the highest.
In contrast, harvesting trees for wood production received the least support (but was still supported by half of the interviewees), with 20% rejecting this rationale.People below the age of 25 expressed the most concern about wood production.
The general rates of acceptance remained similar before and during the first lockdown (p value = 0.988).However, the tolerance for sanitation cuts rose during the lockdown, while the understanding for wood production decreased.There was also a rise of people with no outspoken opinion on the issue of tree felling, except for sanitation cuts (Table 6).Nonetheless, forestry operations, notably logging, do frustrate some forest visitors (Figure 6).The table below displays responses to the open question regarding all frustrations people experience during their forest visits.The main frustrations people mentioned were people using road bikes.'Traffic', the second category often mentioned, mainly referred to forestry vehicles, since no other motorised traffic is permitted in the Kottenforst.
The booming popularity of forest outings during the first lockdown led to an increased dissatisfaction with the behaviour of other forest users.Crowdedness appeared as a totally novel source of annoyance.Complaints about other recreationists' lack of respect dramatically increased compared to the situation before the lockdown.Cyclists remained the main source of frustration, notably road cyclists, but during the first lockdown also increasingly mountain bikers and people on electric bicycles.

Evolution of forest visits
The impact of the COVID-19 related lockdown on the evolution of visitor numbers is clearly illustrated by the counting.During the first lockdown, which started in spring 2020, forest visits increased substantially, before gradually returning to pre-COVID-19 levels after summer.While the lockdown is presumably the main driving factor for the sudden increase in visitor numbers, other factors, such as the sunny spring, cannot be excluded.However, as both springs and summers of 2019 and 2020 had significantly above-average sunshine in the Bonn region, this factor alone is unlikely to explain the visitor boom.The fact that people suddenly had time to spare and that most other leisure options were closed or forbidden were the main pragmatic reasons to explore the outdoors.In line with another study, the vicinity of the forest played a role in its selection as a place for leisure (cf. Beckmann-Wübbelt et al. 2021).With social contact being limited in many situations, a large number of people seemed to have used the forest as a safe place to meet up with friends and family and to maintain social bonds.Similar observations were made in a case study by Weinbrenner et al. (2021)    forest and city park visits in Oslo, Norway, even adjusting for other variables such as the weather.This drastic sudden increase can almost certainly be attributed to the lockdown (cf.also McGinlay et al. 2020;Venter et al. 2020).
Equally drastic as the change in absolute visitor numbers was the change in forest use over the day, clearly reflecting the decrease in commuters and a more intensive use of the forest as a destination for afternoon outings, alone or in a group.This finding seems to contrast with what Ugolini et al. (2020) found during an online survey in six countries, where a decrease in leisure activities but a relative increase of 'necessary' activities, such as walking the dog, was found.This difference can probably be attributed to the stricter conditions for venturing outside in the countries studied (Croatia, Israel, Italy, Lithuania, Slovenia and Spain) compared to Germany, as in the latter country forest visits were not in any way restricted.Ugolini et al. (2020) also found a longing for being outdoors and meeting other people in the six countries covered by their study, which the respondents in our study could put into practice.In contrast, our study indicates that the increased visitor frequencies resulted in some perceptions of crowdedness of the forest, at least by some of the visitors.
The question remains to what extent changing visitor patterns are permanent.One-third of the respondents declared that their frequency of forest visits had increased -less than one would expect based on the counts -but only 13% of all the respondents said they were planning to return more often to the forest even after the end of the pandemic.The quantitative evidence from the most recent counts are ambiguous there -visitor numbers have decreased compared to the lockdown period, but are still a bit higher than in the year before the pandemic.This may possibly indicate that some of the visitors maintained their new habits, especially assuming an 'overcompensation' by others who may have replaced forest visits with different, newly permitted activitiesa phenomenon that may explain the low visitor numbers in the summer of 2020 when the first lockdown was lifted.
Overall, the changing visitor patterns seem to indicate not only the impact of external factors, such as schools or offices being opened or closed, but also the psychological factors.First, there is the effect of habituation: people who get used to a certain situation may fall back on their old behaviour when the conditions characterising the new situation change back to the previous situation.Initially, the novelty of the virus and the projected two-month lockdown may have spurred a sense of urgency leading to drastically altered behaviour, a pattern that can also be observed with other 'new rituals' such as the daily applause for health care workers or the spike in volunteer work.Then, as life returned back to normal, the visit numbers have also stabilised, though still at a slightly higher level than previously.
The visitor profile was characterised mainly by a lack of young people.The share of people under the age of 40 was barely 20%.The average age was 52, compared with 42 in the city of Bonn.The reasons for the low popularity of the forest among young people are unclear but may be connected to a general lower tendency towards outdoor recreation in millennials (cf.Mehlhaf 2019).During the lockdown, the foresters working in the Kottenforst reported a tangible increase in visitors going to the forest in the company of family, notably children, as well as an increase in the share of young respondents.Nonetheless, the average age of the interviewees went up, despite the observations of the interviewers and the foresters that more young people were discovering the forest.This can be explained by several factors, one being that when interviewing families, only the age of the spokesperson of the group was recorded, leaving out children from the calculations.
According to our interviews, the main drivers for increased (or decreased) forest visits were not directly related to the virus (eg.fear of indoor spaces, enhancing the immune system, attempts to avoid other people), but to the effects of the restrictions implemented to limit the infection rate.Closed schools meant that parents had to keep their children busy during the day, and the more flexible working schedule for people who were able to do home office allowed for more frequent outings.The lack of alternative leisure activities was another important driver for forest visits.The role of the virus as an immediate driver for a changed attitude vis-à-vis the forest was found to be marginal.Only one respondent indicated using the forest to cope with the fear of the virus, and another that fresh forest air strengthens the immune system.A general increased concern with health could be detected, however, with more people stressing the importance of exercise than before the lockdown.These findings suggest that keeping the forest open to visitors was the right policy decision, and that closing them off would have had negative consequences for public health (see also Ugolini et al. 2020).
This case study does generally confirm the importance of peri-urban forests for recreation and wellbeing and stresses the forest's importance for people faced with unforeseen alterations of their daily routines, mirroring the findings of Weinbrenner et al. (2021) in Freiburg, or Yiyang et al. (2021) who describe the cushioning effect of urban green spaces on the decrease in leisure-time physical activity in Asian cities, by mitigating the decrease of other leisure activities that had become obsolete.The effect was higher in people living close to green spaces, highlighting the importance of accessible urban greenery.

Associations with the Kottenforst
Our study largely confirms previous research when it comes to the factor that determine forest recreation.The fact that the vast majority of visitors live right next to the forest confirms the study of Meyer et al. (2019), who stress the importance of vicinity and spontaneity when people plan a forest visit.Spiritual and escapist motives (cf.Rametsteiner and Kraxner 2003) were prominent in the answers of over half of the respondents, mainly when referring to aesthetic elements and the emotional connection they have with the forest.As expected from a largely urban sample group, the value attributed to the productive function of the forest was lower compared to recreational benefits (Upton et al. 2014;Lorenz and Elsasser 2018), and peace and quiet were the main motivations to visit the forest for 46% of the interviewees.In accordance with other studies (cf.Pietilä et al. 2015;Roviello et al. 2021), numerous respondents expected mental and/or physical health benefits from their forest visit, connected to the stated motivations of physical exercise and fresh air.Regarding perceived stand beauty and naturalness, the overall opinion was a dislike of visual traces of human interventions, also found in other studies (cf. Tindall 2003;Ciesielski and Stereńczak 2018), illustrated by the lower values attributed to monocultures and the aversion for machinery in the forest.The preference for the idea of perceived wilderness, also found by among other in Ranacher et al. (2017), was demonstrated by the higher aesthetic values attributed to those stands that were perceived and described as wild or unmanaged.
The general associations of visitors with the forest remained remarkably stable when comparing the interviews during the lockdown with those before the pandemic.Tranquillity remained the main association people had with the forest, but did lose some importance compared to before the lockdown, presumably because of the increased crowdedness.During the lockdown, the colour green gained importance as a predictor for perceived stand beauty and naturalness.This could partly be explained by the arrival of occasional visitors for whom the contrast to the city may be more poignant.Another factor is that all the lockdown interviews were carried out during the spring, a period characterised by the return of green to the landscape.It is not surprising that this association was less prevalent in interviews done in autumn or winter, before the lockdown.The increased focus on tree health could be attributed to two factors.The first and most obvious one would be the progressing dieback of the bark-beetle-infested spruce after yet another dry summer, which was extensively covered by the media.Secondly, the general focus on health during the lockdown may be reflected in people more consciously observing health issues in their surroundings.

The impact of forest management on recreation
The quest for forest characteristics that contribute to the visitors' appreciation of the forest did yield some clear results, which may be of value for the forest management strategy.
The clear preferences for uneven-aged stands resulting from selective cutting regimes speaks to the societal acceptance of the management style that has been practiced in the Kottenforst for decades.Newly established stands were also perceived as beautiful and natural, but not one respondent expressed awareness of the clearcut origin of these stands.The stands that were perceived as the most beautiful and most natural looking belonged to the Habitat Type 9160.This forest type, dominated by oak and hornbeam, is the result of centuries of human management including past agroforestry practices and would possibly turn into a beechdominated forest type without regular human interventions.The preference for this forest type, which formed over centuries of human management, implies that while people dislike seeing human activity in the forest, they do like some of its long-term outcomes, something that was also observed by Konczal et al. (2022).The previously mentioned increased visibility and awareness of spruce dieback could explain the increased acceptance of felling sick trees, while the general acceptance of tree cutting was lower.During the lockdown, there were also more people who indicated being neutral or having no opinion on tree felling, possibly because of a rise in novel, occasional forest visitors.
While road cyclists remained the top frustration during the lockdown (even in the eyes of other cyclists), new nuisances appeared that are probably related to the increase in visitor numbers.Disrespectful behaviour and increased waste were prevalent sources of irritation during the lockdown.While the associations with the forest remained largely the same, the reasons for going there and the use of the space did change to a certain degree.The increased use of the forest as a space for encounters suggests what Weinbrenner et al. (2021) call 'spatial appropriation' of the forest.According to this study, the forest was increasingly seen as an extension of the living room, and used for various -social-activities, not restricted to forest walks.With more people appropriating the forest for various activities, crowdedness became a novel concern.Crowdedness is a topic that has not received much attention in forest research in Europe so far (cf.Arnberger and Mann 2008), but is becoming important for land use planning in densely populated areas.In our study, the visitor's sense of forest crowding was a key observation made by several interviewees, which concords with the findings of McGinlay et al. (2020) but which was not found in the study of Weinbrenner et al. (2021).This may be related to the initial outset of the interviews; in our study, people were specifically asked about their frustrations.Future research should strengthen insight in forest visitor perceptions towards crowding, and in how land use planning, forest design and nature management can support higher visitor numbers in nature and green spaces.Keeping track of visitor flows may also become increasingly helpful in urbanised regions in order to plan the timing of forest interventions and to plan the workload related to the reception of different visitor types throughout the day, the week and the season.
These exceptional circumstances have led to observations that may be temporary but do hint at the bigger picture of what urban populations expect from the forests in their surroundings.Given the growing societal interest in forests, conducting similar studies in the future may assist in aligning the preferences and priorities of forest visitors and forest managers.The integrated forest management approach practiced in the Kottenforst seems to have proven to be quite successful in catering to the diverse expectations of the visitors while still ensuring revenue through wood production.Yet, transparent and efficient communication will be critical for the success of the concept in the future (Winkel and Spellmann 2019).

Methodological limitations
While we think our study's methodology has considerable value, especially through the combination of a longitudinal visitor counting with an assessment of visitor perceptions and motivations before and during the lockdown, we also wish to acknowledge its limitations.
First, this research was the continuation of a study that started in early 2019 as an inquiry of visitor perception of the management of the Kottenforst, specifically in relation to the restoration of vulnerable habitats, notably oak-hornbeam dominated forest (habitat type 9160).It was the first study of forest recreation ever in this region, and therefore the initial approach was broad and covered various aspects of forest management and recreation.When the pandemic started, we were among the first to publish quantitative data on the surge in forest visits.We then decided to expand on the ongoing study with follow-up interviews.This explains the exploratory approach of the paper, guided by a suddenly changed reality.
When interpreting the findings in this paper it needs to be noted that the number of interviewees was limited, especially for the second round of interviews during the lockdown.This resulted in rather small absolute numbers of respondents for some questions where only a few respondents answered positively.For instance, this was the case for the follow-up question during the lockdown about reasons to go to the forest more often, which only applied to 18 people.This is a clear limitation of our data.At the same time the fact that we conducted the interviews with forest visitors during the lockdown, and not 'ex post' asking them afterwards about their motivation to visit the forest, can be seen as a strength of the study.To conclude, while we are confident that this study adds interesting evidence to the forest recreation literature under the circumstances of a major public health (or public life) crisis, it is important to consider the exploratory case study character of our work.

Conclusion
This exploratory study adds one example to the increasing body of literature demonstrating that strict measures taken up in response to the public health crisis caused by COVID-19 resulted in many and partially unforeseen effects in different fields of society.The obvious role that forests have played during this crisis has arguably made them part of the critical infrastructure society depends on, especially also in the time of a crisis.The importance of forest recreation as the primary ecosystem service in the Kottenforst became even clearer during the pandemic.However, managing the forest for other ecosystem services has led to the image that is now enjoyed by the visitors.In order to prepare for potential new waves of visitors, adapted forest management (planning) may however still be needed.
Urban forests have long since served as a place where people can escape the hustle of the cities.The COVID-19 pandemic, with its associated psychological impacts and government-imposed restrictions, has demonstrated this more clearly than ever before.Forests can act as a decompression valve during times of increased pressure, and as a green space that is theoretically accessible to everyone.In an urban setting, people mainly go to the forest to find peace and quiet and to enjoy the scenery.However, as this study shows, these exact pursuits became increasingly difficult when visitor numbers increased.Because of the rise in visitors, crowdedness turned out to be a concern in an unprecedented dimension.Forest crowding conflicted with people's two main motivations -tranquillity and beauty -and is thus a crucial factor when planning recreational forest use.A new sense of spatial appropriation of the forest may also play a role.While finding tranquillity was the prime motivation of most people to go to the forest -both before and during the lockdown -an increase of people going to the forest with friends and family was observed during the lockdown.The lack of other social activities highlighted that forests offer solace, not only to those looking for solitude but also those looking for company.This in turn can potentially lead to conflicts between those searching peace and those searching entertainment.
The new cohort of forest enthusiasts, partly younger people, may have a different perception of forest management practices.The decreased acceptance of wood harvesting and the larger number of observers pose a challenge for forest managers, who are trained to optimise wood production but not so much to optimise recreation opportunities.
The increased popularity of the Kottenforst may require novel communication strategies aimed at enhancing the mutual understanding of different forest users.Integrated forest management has the potential to do so but needs to be flexible and adaptive to social needs, which can, as the lockdown has demonstrated, change abruptly.

Figure 1 .
Figure 1.Interview locations and location of the infrared counter, as well as the location of the Kottenforst in the larger Cologne-Bonn region.The delineated area lies within the Natura 2000 network.The figure was generated based on the map used in the forest management plan.(Source: Wald und Holz NRW).

Figure 2 .
Figure 2. Timeline of weekly visit counts since the start of the counting.

Figure 3 .
Figure 3. Hourly visitor counts during the days of the week, compared between four periods.Top left: pre-lockdown -spring.Top right: first lockdown -spring.Bottom left: pre-lockdown -winter.Bottom right: second lockdown -winter.

Figure 5 .
Figure 5. Answers to the question 'Which three words do you associate with the Kottenforst?' Before and during the first lockdown, in percentage of interviewees.(N before lockdown = 292, N during lockdown = 53).

Figure 4 .
Figure 4. Overview of individual self-declared changes in visit frequency during and after the lockdown (N total = 53).

Figure 6 .
Figure 6.Overview of the interviewees' main frustrations regarding their forest visits, before and during the lockdown.(N before lockdown = 292, N during lockdown = 53).The overall difference between the 2 periods is not statistically significant (p = 0.711).

Table 1 .
Overview of age and gender of the respondents, before and during the first lockdown.

Table 2 .
Overview of the keywords per coding category, before and during the first lockdown.

Table 3 .
Overview Wilhelm 2013)assessment, shown as average scores on perceived beauty and naturalness, combining the two interview periods (N = 345).The age classes are divided into establishment (1 -young, successional forest), qualification (2 -still rather young often dense forest), dimensioning (3 -middle age forest) and maturity (4 -old forest with larger dimension trees) phase (cf.Wilhelm 2013).B stands for broadleaf, C for coniferous and M for mixed stand.Respondents rated beauty and naturalness on a 5-point Likert scale 5.For information the typical oak-hornbeam forests of habitat type 9160, which were the focus of the LIFE+ project, are marked as such.

Table 4 .
Indications for the perceived beauty of the stand where people were interviewed, before and during lockdown, in percentage of interviewees as response to the open question to support their score with arguments (N before lockdown = 292, N during lockdown = 53).The overall difference between the 2 periods is not statistically significant (p = 0.274).
in the German city of Freiburg.The recorded changes during the lockdown are largely in line with what has been observed in other studies in different countries.Using mobile tracking data, Venter et al. (2020) found a 291% increase in

Table 6 .
Detailed overview of acceptance of wood felling before and during lockdown, differentiating according to a main rationale for why trees are felled, in percentage of interviewees.(N before lockdown = 292, N during lockdown = 53).

Table 5 .
Indications for the perceived naturalness of the stand where people were interviewed, before and during lockdown, in percentage of interviewees as response to the open question to support their score with arguments (N before lockdown = 292, N during lockdown = 53).The overall difference between the 2 periods is not statistically significant (p = 0.189).