Evidence of COVID-19 pandemic influence on well-being produced by urban gardening: a before-after study

A growing majority of people live in cities, where dense, interconnected systems heighten vulnerability to disturbances such as pandemics and climate change. Strengthening individual and community resilience—through internal resources, social support, social networks, social capital, and collective learning—is therefore critical. Urban green spaces and green commons (e.g., parks, community and allotment gardens) are key supports of resilience by providing ecosystem services that promote physical and mental health, biodiversity, social cohesion, and social networks. COVID-19 altered social, economic, and political settings, likely changing attitudes toward and outcomes from urban green spaces. While many reported increased reliance on nature and green spaces during restrictions, access varied with socioeconomic factors, and some reduced visits due to crowding or infection concerns. Globally, a "gardening boom" was observed, with gardening offering mental health benefits and community resilience, but also new barriers (e.g., supply shortages, crowding). Yet, evidence is limited on how gardening benefits changed compared to pre-pandemic and whether early increases persisted. This study examines urban box gardening on public land in Turku, Finland, before (2019), during the first COVID summer (2020), and after restrictions loosened (2021). Using the Social-Ecological Systems framework, the study asks whether the importance assigned to gardening increased due to the pandemic, and whether ecological outcomes (cultivation success) and self-perceived outcomes changed in 2020 and 2021 relative to 2019. Given reports of more time for gardening and higher perceived importance, the authors expected increases in both ecological and self-perceived outcomes during the pandemic.

Prior work shows urban green spaces bolster resilience through ecosystem services, enhancing mental and physical health, biodiversity, and social cohesion. During COVID-19, many turned to nature for coping, with increased visits to green areas where permitted, and higher stated importance of urban green spaces, especially among those without private gardens. However, access was uneven due to socioeconomic and policy factors; some avoided crowded green areas. A global surge in gardening was reported early in the pandemic, with gardening linked to stress relief, positive emotions, and mental resilience, while also facing barriers such as supply shortages and crowding. Urban gardening can strengthen community resilience by building social capital, informal networks, and collective learning, and it can buffer food insecurity by providing access to fresh produce. Despite these documented benefits, gaps remain on how the benefits from urban gardening changed relative to pre-pandemic conditions and whether initial increases persisted in the longer term after lockdowns eased, and whether attitude changes translated into actions.

Setting and program: The study was conducted in Turku, Southern Finland, over summers 2019, 2020, and 2021 within a municipal program that provides 1 m² garden boxes and soil on public land to residents. Gardeners self-organize groups, choose locations on public land, and manage cultivations. Box counts and locations: 2019: 698 boxes at 245 locations; 2020: 674 boxes at 243 locations; 2021: 762 boxes at 297 locations. Applications/new boxes: 2019: 116 applications, 70 new boxes; 2020: 46 applications, 40 new boxes; 2021: 171 applications, 110 new boxes. Data collection framework: Guided by the Social-Ecological Systems framework to identify variables at the resource user level. Two data streams were collected annually: (1) field inventories to measure ecological outcomes (what grows in boxes), and (2) electronic questionnaires to measure self-perceived outcomes and gardener/group characteristics. Field inventories: Conducted three times each summer (~June, July, September), capturing Finland’s short growing season. For each box, assess: cultivated species identity; number of cultivated individuals; quality of each species; area cover of cultivations; share of weed cover; and estimate economic value of cultivations per box (following Tuominen et al. procedure). Record site privacy (secluded vs public) and sun exposure (shade). Questionnaire: Two online surveys (before and after season; analyses use after-season responses). Self-perceived outcomes: gardeners rated 14 benefits (fresh vegetables, physical exercise, mental relaxation, beautifying the area, educating children, self-sufficiency, new friends, community feeling, quality time with friends and family, knowhow, happiness, belonging to a social movement, nature experiences, creating biodiversity) as not received / received a little / received a lot. Group/activity covariates: group size, presence of family members in the group, frequency of group meetings, number of other gardeners at the location, starting year (new vs experienced), number of boxes, privacy, shade. COVID-specific questions (2020–2021): changes in gardening, attitude toward gardening, and economic situation due to the pandemic. Linking field and survey data: respondents provided their gardening group name (used by the city and posted on boxes), enabling linkage; accidental duplicates removed. Sample sizes and response rates: For analyses, after linking to field data: n = 216 (2019: 67; 2020: 65; 2021: 84). Overall after-season survey respondents across years: 234. Response rates (after-season): 2019: 73/256 (28.5%); 2020: 69/258 (26.7%); 2021: 92/308 (29.9%). Across all years, 420 respondents answered out of 822 gardening locations (51.1% any-survey rate); 28.4% after-season rate. Non-respondent considerations: in 2021, ~25% of non-respondents had quit gardening, potentially biasing ecological comparisons; respondents generally showed higher cultivation success than non-respondents. Socio-demographics (available 2020–2021): about 40% graduate degree, 25% undergraduate, 30% secondary; 2021 age: ~40% aged 30–39 with fairly even distribution up to age 70; 78% female. New gardeners share in analysis dataset: 2019: 41.8%; 2020: 21.5%; 2021: 53.6%; overall 40.3%. New gardeners outperformed experienced gardeners on several ecological metrics; starting year included in models. Statistical analyses: (1) Binomial sign tests (two-tailed) assessed whether self-reported changes in attitudes and economic situation (2020–2021) were more positive or negative than expected by chance. (2) Principal Component Analyses (PCAs) reduced dimensionality separately for ecological outcomes (6 variables: produce quality, species per box, cultivated area per box, individuals per box, economic value per box, weed cover share) and self-perceived outcomes (14 benefit variables). PCA scaling standardized variances; components with eigenvalues >1 and sensible loadings were retained. (3) General Linear Mixed Models (GLMMs) tested year differences (2019 vs 2020 vs 2021) in PCA component scores. Fixed effects: year plus covariates (group size, family members, meeting frequency, number of other gardeners, starting year, number of boxes, privacy, shade). Random effect: gardener ID to account for repeated participation. Assumptions were checked and acceptable. Analyses performed in R with FactoMineR (PCA) and lme4 (GLMM).

Pandemic-related attitudes and economy: In 2020, 28/139 respondents (20%) reported attitude change toward gardening due to COVID-19 (27 positive, 1 negative); binomial test p < 0.001. In 2021, 44/145 (30%) reported attitude change (42 positive, 2 negative); p < 0.001. Thus, among those reporting change, attitudes were significantly more positive in both years, with slightly more positive changes in 2021. Economic situation changes were reported by ~20%: 2020: 30/139 (25 negative, 5 positive), p < 0.001; 2021: 28/155 (21 negative, 7 positive), p = 0.013. A panel of 54 respondents answering both years showed qualitatively similar results. Ecological outcomes (PCA): Two components (eigenvalues >1) explained ~55% of variance (PC1 ~38%, PC2 ~17%). PC1 “Overall ecological outcome” loaded positively on species, cultivated area, economic value, individuals, and negatively on weed cover. PC2 “Simpler strategy for quality” loaded positively on quality and negatively on species and individuals. Year effects: No significant differences between 2019 and 2020 (df=90.9, t=0.181, p=0.857) nor 2019 and 2021 (df=108.2, t=0.114, p=0.909) for PC1; no difference between 2020 and 2021 (t=0.054, p=0.957). Predictors of higher PC1: fewer years gardening (df=194.2, t=5.600, p<0.001) and less sunny locations (df=202.8, t=−2.137, p=0.034). For PC2, significant differences: 2019 vs 2020 (df=143.01, t=4.646, p<0.001) and 2019 vs 2021 (df=163.44, t=2.754, p=0.007), indicating in 2020–2021 gardeners produced higher-quality output with fewer species/individuals than in 2019; 2020 vs 2021 not significant (t=1.661, p=0.099). Self-perceived outcomes (PCA): Three components retained (PC1 ~34%, PC2 ~9%, PC3 ~9%). PC1 “Overall self-perceived outcome” reflected receiving more of all listed benefits at higher scores. PC2 “High social, low practical outcomes” reflected greater social/non-tangible benefits (beautification, community) versus practical cultivation benefits (fresh vegetables, self-sufficiency). PC3 “High family, low individual outcomes” reflected family-related benefits versus individual-oriented benefits. Year effects: PC1 was significantly lower in 2021 than 2019 (df=130.2, t=−2.468, p=0.015); 2020 vs 2019 showed a non-significant decreasing trend (df=108.4, t=−1.021, p=0.309); 2020 vs 2021 not significant (t=1.472, p=0.143). Predictors of higher PC1: more frequent meetings (df=201.5, t=2.074, p=0.039) and sunnier locations (df=204.1, t=2.191, p=0.030). For PC2, 2021 was higher than 2019 (df=168.29, t=2.179, p=0.031), indicating a shift toward social over practical benefits; 2020 vs 2019 not significant (t=0.820, p=0.413); 2020 vs 2021 not significant (p=0.173). Predictors of higher PC2: larger group size (df=202.55, t=3.427, p<0.001) and fewer years gardening (df=191.62, t=2.533, p=0.012). For PC3, no significant year differences (2020: p=0.707; 2021: p=0.640). Predictors of higher PC3: gardening with family members (df=203.39, t=2.171, p=0.031), more frequent meetings (df=202.10, t=3.768, p<0.001), and more public locations (df=203.87, t=−1.997, p=0.047) associated with higher family-oriented benefits. Synthesis: Despite more positive attitudes and stable ecological outcomes, gardeners reported fewer overall self-perceived benefits in 2021 than pre-pandemic, with a shift toward social over practical benefits. Ecological strategy shifted toward cultivating fewer species/individuals with higher quality during the pandemic, without changing overall ecological outputs.

The study asked whether COVID-19 altered attitudes toward urban box gardening and whether ecological and self-perceived outcomes changed relative to pre-pandemic conditions. Results show that when attitudes changed, they overwhelmingly shifted positively, underscoring the role of urban gardening in individual resilience during crises. However, contrary to expectations, overall self-perceived benefits decreased by 2021 relative to 2019, suggesting the broader negative effects of the pandemic on well-being extended into gardening-related benefits. Potential mechanisms include concerns about distancing in public locations, general declines in mental and physical health during the pandemic, and elevated expectations as gardening gained importance and time availability increased. At the same time, self-perceived benefits shifted toward social, non-material outcomes, especially in larger groups and among less experienced gardeners, indicating strengthened social motivations and potential contributions to community resilience. Ecological outcomes remained stable across years, suggesting sustained motivation and capacity to maintain cultivation success and local biodiversity despite societal disturbance. A shift toward a simpler cultivation strategy (fewer species/individuals with higher quality) emerged during the pandemic, potentially reflecting increased focus, innovation in techniques, or supply constraints. The decoupling between self-perceived and ecological outcomes aligns with prior evidence and implies non-material values may be more sensitive to societal stressors than material outputs. Collectively, increased positive attitudes, sustained ecological outputs, and enduring changes in perceived benefits and cultivation strategy point to adaptive, possibly transformative, change in this social-ecological system. Programs enabling collective action and commons-like organization may act as hubs for social and sustainable transformation, strengthening community resilience.

Urban box gardening in Turku supported individual and community resilience during COVID-19, as evidenced by more positive attitudes and sustained ecological outcomes. Yet, gardeners reported fewer overall self-perceived benefits by 2021 compared to 2019 and a shift toward social over practical benefits, indicating that the pandemic’s negative well-being impacts extended into the gardening domain. Ecologically, a pandemic-era shift toward a simpler, higher-quality cultivation strategy occurred without changing overall ecological outcomes. These findings demonstrate that outcomes from small-scale social-ecological systems are sensitive to broader societal changes and may undergo adaptive or transformative shifts. Policy and practice should prioritize maintaining open, safe access to urban green spaces during crises, and strengthen programs toward urban green commons to enhance social networks, collective learning, and empowerment. Participatory planning can help recognize diverse needs and secure key benefits during disturbances. Future work should verify longer-term transformational change and ensure equitable access and benefits across socioeconomic groups.

The sample comprises urban box gardeners and is not representative of the general population, limiting generalizability. Socioeconomic attributes were not collected in 2019 and only partially in 2020–2021, constraining analyses of equity and access. Objective measures of well-being were not collected, preventing direct assessment of health impacts or comparison with non-gardeners. Response bias is possible; respondents generally had more successful cultivations than non-respondents, and a substantial fraction of non-respondents had ceased gardening, potentially biasing ecological comparisons. The mapping between individual authors and affiliations in the provided text is not explicit. Finally, results are context-specific to Turku’s public, small-scale box gardening program.