Bottom-up and top-down drivers influence urbanization effects on insect herbivory in oaks
Data files
Oct 24, 2025 version files 227.71 KB
-
clima.csv
137.30 KB
-
data.csv
31.77 KB
-
nutrients.csv
3.88 KB
-
phenolics.csv
17.94 KB
-
README.md
4.87 KB
-
URBANIZATION_DATA_ANALYSIS_OIKOS.Rmd
31.95 KB
Abstract
Urban forests provide essential ecosystem services. Herbivory is a widespread biotic interaction that shapes ecosystem functions, which underpin these services. Urbanization disrupts plant-herbivore interactions by altering plant traits (bottom-up factors) or by changing the abundance of herbivore natural enemies (top-down control), potentially threatening the services provided by urban forests. Disentangling these drivers of herbivory is crucial for designing and managing urban forests to enhance resilience. We examined insect leaf herbivory on Quercus robur trees in urban and rural forest stands across 13 European cities (N = 104 trees). To assess the effects of top-down forces on herbivory, we excluded vertebrate (e.g., birds, bats), invertebrate (e.g., ants), or both groups of predators from branches on each tree. We then measured insect damage on the different treatments. We also assessed bottom-up drivers by measuring leaf traits, specifically nutrients and phenolic compounds, and tested for correlations with leaf damage. Additionally, we recorded temperature within stands, an abiotic factor that may modulate both top-down and bottom-up forces on herbivory. Herbivory was 24% lower on urban trees compared to rural trees. Excluding vertebrate predators increased herbivory by 40%, with effects being stronger in urban stands. Additionally, urban trees had higher leaf quality, with higher nutrient and lower phenolic concentrations; however, the measured traits did not correlate with herbivory. Finally, temperature was positively associated with urbanization and correlated with predation (positively), but not with herbivory. Temperature also did not mediate the bottom-up or top-down effects of urbanization on herbivory. Urbanization affects herbivory through both bottom-up and top-down processes, independent of local temperature. Despite stronger predator effects and better leaf quality, urban trees experience lower herbivory, suggesting that unmeasured factors, such as changes in herbivore behaviour or community structure, may play an important role. Further studies are needed to deepen our understanding and inform urban forest management.
Dataset DOI: 10.5061/dryad.s7h44j1mr
Description of the data and file structure
In this study, we examined insect leaf herbivory on Quercus robur trees in urban and rural forest stands across 13 European cities (N = 104 trees). To assess the effects of top-down forces on herbivory, we excluded vertebrate (e.g., birds, bats), invertebrate (e.g., ants), or both groups of predators from branches on each tree. We then measured insect damage on the different treatments. We also assessed bottom-up drivers by measuring leaf traits, specifically nutrients and phenolic compounds, and tested for correlations with leaf damage. Additionally, we recorded temperature within stands, an abiotic factor that may modulate both top-down and bottom-up forces on herbivory.
Files and variables
File: clima.csv
Description: At the beginning of the growing season, we installed an Elitech RC-5 data logger (Elitech, London, UK) 1.5 meters above the ground on a tree in each stand to quantify air temperature. This file contains the daily mean temperature for each stand in each city (26 stands across 13 cities)
Variables
- date: Day when the data were collected
- meantemp: Daily mean temperature (°C) of the stand
- Urb: Urban or rural stands within each site
- Site: City (among 13 European cities)
File: data.csv
Description: Chewing herbivory data estimated by calculating the percentage of leaf area removed from each leaf by using the BioLeaf – Foliar AnalysisTM mobile app.
Missing values:
- "n/a" indicates cases where no measurement was taken because insufficient leaf material was available for analysis in that treatment.
Variables
- ID: Unique identifier for each sample
- Site: City (among 13 European cities)
- T: Branch treatment (Ant-excluded, Bird-excluded, Both, Control)
- Urb: Urban or rural stands within each site
- Tree: Tree number (4 trees per stand)
- L1-L10: Individual leaf herbivory values (% leaf area removed)
- herbivory: Mean leaf herbivory per branch
File: nutrients.csv
Description: Leaf nitrogen (N) and phosphorus (P) concentrations. Nitrogen was quantified using the indophenol blue method, and phosphorus using the molybdenum blue method, with absorbance measured at 650 nm and 700 nm, respectively, using a Biorad 650 microplate reader (Walinga et al., 1995). Both concentrations were expressed in mg g⁻¹ dry tissue.
Variables
- ID: Sample number
- Site: City (among 13 European cities)
- T: Branch treatment (Ant-excluded, Bird-excluded, Both, Control)
- Urb: Urban or rural stands within each site
- Tree: Tree number (4 trees per stand)
- P: Phosphorus concentration (mg g⁻¹)
- N: Nitrogen concentration (mg g⁻¹)
File: phenolics.csv
Description: For each tree, we selected four leaves (out of the ten originally sampled) from control branches showing little or no herbivore damage for measuring phenolic compounds by UHPLC equipped with a UV/VIS photodiode array detector. Analysis focused on: flavonoids, phenolic acids, hydrolysable tannins, condensed tannins, and total phenolics. All concentrations were expressed in mg g⁻¹ of dry tissue.
Missing values:
- "n/a" indicates that the measurement was below the detection limit of the instrument, and therefore, no quantitative value could be reported for the specific compound.
Variables
- ID: Sample number
- Site: City (among 13 European cities)
- T: Branch treatment (Ant-excluded, Bird-excluded, Both, Control)
- Urb: Urban or rural stands within each site
- Tree: Tree number (4 trees per stand)
- F1-F8: Individual flavonoid compound concentration (mg g⁻¹)
- PA1-PA3: Individual phenolic acid concentration (mg g⁻¹)
- CT: Condensed tannin concentration (mg g⁻¹)
- HT1-HT5: Hydrolysable tannin concentration (mg g⁻¹)
- flav: Total flavonoid concentration (mg g⁻¹)
- phen_acids: Total phenolic acid concentration (mg g⁻¹)
- condensed_tannins: Total condensed tannin concentration (mg g⁻¹)
- hydrolysable_tannins: Total hydrolysable tannin concentration (mg g⁻¹)
- total: Total phenolic concentration (mg g⁻¹)
File: URBANIZATION_DATA_ANALYSIS_OIKOS.Rmd
Description: R Markdown document containing the code used for data analysis and figure generation in this study.
Code/software
All statistical procedures were carried out in R version 4.2.1 (R Core Team, 2024). The R Markdown file "URBANIZATION_DATA_ANALYSIS_OIKOS.Rmd" contains all the necessary code to replicate the results of the study, including loading the data sets, performing data manipulations, creating subsets for analysis, and running all the models.