Abstract:

Arboricultural research focusing on transport land use was lacking in Hong Kong. Some highway slopes were registered in the Systematic Identification of Maintenance Responsibility of Slopes in the Territory (SIMAR), abbreviated as SIMAR slopes. We aimed to analyze patterns in the structure and species composition of the tree stock along a highway in Hong Kong and examined how a slope registration system could help explain the characteristics of urban forests. The 53 slopes and 52 verges along San Tin Highway, Hong Kong were randomly selected. The trees on each slope and verge were collectively sampled as a tree stand. Six variables, namely tree abundance, species richness, maximum tree height, Shannon–Wiener diversity, Simpson’s dominance, and Pielou’s evenness were measured for each stand. In addition, a limited visual tree risk assessment was performed. The 7,209 trees in 23 species were recorded. Species richness was low, ranging from one to eight species per stand. SIMAR and non-SIMAR slopes were compared. SIMAR slopes had significantly higher species richness, diversity, evenness but lower dominance, with mean difference of 1.41 species, 0.17, 0.17 and − 0.28 respectively. SIMAR slopes were associated with lower tree risk rating. When training regression models, boosting as an ensemble method arbitrarily raised the explanatory power and the predictive accuracy of some models. Slope height, length, angle and area could be significant predictors of the biodiversity-related variables. Future research can sample more habitat characteristics related to the structure and species composition of slopes and verges which were important components of urban forestry.

Key words: Highway greenery, Roadside tree management, Tree risk assessment, Monospecific urban forest, Slope geophysical environment, Urban forests