Restoration of urban forest improves quality of life for city residents and is important for boosting native biodiversity. However, the scientific knowledge required to inform successful restoration is largely lacking for urban forest ecosystems, which differ from rural forests. Here we review two decades of urban forest restoration research in Hamilton, New Zealand and summarise key findings across seven major ecological topics: i. species traits, filters and thresholds, ii. species richness and target ecosystems, iii. tree regeneration, iv. seed banks and seed rain, v. seed predation, vi. enrichment planting and vii. restored forest function. We then discuss general urban restoration principles that increase the efficacy of urban restoration efforts including restore to a minimum of 10% indigenous cover in cities, develop a step‐wise restoration plan and prioritise partner engagement. The purpose of this review is to aid urban forest restoration across New Zealand and globally.

Forest restoration has potential to recover degraded ecosystem functions in disturbed environments. Decomposition and denitrification are two critical functions involved in forest nutrient cycling that are often compromised in degraded ecosystems. As forest canopy structure develops following initial plantings, it may indirectly impact ecosystem functions by altering abiotic conditions. It is likely, however, that there are other abiotic factors that affect decomposition and denitrification that are unrelated to forest canopy structure. Here, we aimed to determine whether forest canopy openness, topography, and soil sand content would affect litter decomposition and denitrification by regulating the microclimate, the herbaceous plant layer, soil chemistry, and soil moisture. Research occurred in restored native temperate rainforest patches in two New Zealand cities. Urban forests are an excellent context for measuring impact of canopy restoration on ecosystem properties such as microclimate due to the extreme swings in city conditions (e.g., urban heat island). Decomposition rates were determined using leaf litter bags and denitrification rates through denitrification enzyme activity assays. We used structural equation modeling to quantify the direct and indirect drivers of these ecosystem functions. Results indicated that decomposition rates were positively related to soil moisture, relative humidity, and herbaceous plant cover. Interestingly, forest canopy openness indirectly affected decomposition through counteracting forces, meaning greater canopy openness in young forests permitted dense herbaceous plant growth which enhanced decomposition, while less canopy openness in older forests enhanced humidity levels which increased decomposition. Denitrification was negatively related to soil pH and positively related to soil moisture, but these abiotic factors were unrelated to the forest canopy. Discovering drivers of ecosystem functions can improve approaches to the restoration of degraded ecosystems, especially in disturbed urban areas. Identifying counteracting effects on ecosystem functions could improve management by focusing restoration actions on specific drivers to elicit desired changes. Some ecosystem processes, like denitrification, are not affected by forest canopy restoration or management, but are instead driven by edaphic and landscape factors.