Urban soil respiration patterns: An analysis of Quercus robur vs. Quercus rubra in parks and streets of Karlsruhe, Germany

The dynamics of soil respiration (R_s) in natural forest and urban forest ecosystems are already well-researched. However, there is still little research on soil respiration under individual trees in urban areas, particularly regarding the influence of the habitat, different tree species, development stage, and their interplay on soil respiration. In this study, we examined how R_s varies between the two most common urban tree-habitats (park and street), two tree species (Quercus robur L. and Quercus rubra L.), and tree development stages (a three-tier classification system based on the developmental stages of urban trees which is linked to the time since planting). We hypothesized that R_s values would differ between habitats but not between the native Quercus robur and non-native Quercus rubra. Additionally, we expected a decrease in R_s with the advancing development stage. The study involved 60 trees (30 per tree species, 20 per development stage) distributed across Karlsruhe, southwest Germany. We measured R_s using a portable soil gas chamber. Alongside R_s, environmental (site and soil attributes), and dendrometric data were collected. Regression analysis revealed an interaction between tree species and development stage with statistically significant effect on R_s. Quercus robur displayed contrasting R_s patterns in parks (increase) and streets (initial increase, then decrease) with progressing development stage. In contrast, Quercus rubra showed only in streets a clear (negative) trend. While Quercus rubra R_s differed between locations only in the last development stage, Quercus robur showed higher R_s in streets than in parks in the first development stage, with a reversal in the last stage. Additionally, Quercus rubra exhibited overall higher mean R_s values than Q. robur (4.20 ± 0.25 µmol m⁻² s⁻¹ vs. 3.56 ± 0^.18 µmol m⁻² s⁻¹). Soil water content, soil water content × soil temperature interaction, soil pH, and impervious surface cover had also a statistically significant effect on R_s. Parks had higher median soil water content (0.11 ± 0.07 m³m⁻³) and soil pH levels (6 ± 0.74) compared to streets (0.06 ± 0.06 m³m⁻³; 5.5 ± 0), while streets had a greater median of impervious surface beneath the tree canopy (20 ± 29 % vs. 0 ± 0 %). Additionally, lower values for R_s and soil pH, which can also be indicators of soil quality, may suggest poorer soil quality at older sites and street locations. Further investigations could clarify whether this is true and if the varying application of soil amendments by the horticultural department is causing the different trends in R_s rates, as well a_s which other factors further explain the variance in R_s.