Data from: Heat vulnerability and leaf traits affect litter decomposition across an urban heat island gradient in Philadelphia, PA.

Here, we evaluated the effects of heat vulnerability (i.e., a proxy for extreme temperatures), litter quality (i.e., labile versus recalcitrant), and methodology on litter decomposition across an urban heat island in Philadelphia, PA. While the city has actively worked to expand urban tree canopy cover (Philadelphia Parks & Recreation 2023), recent mapping efforts have identified distinctive UHIs throughout the city (The Nature Conservancy 2023). We leveraged these maps to identify four sites that represented a gradient in temperature and Heat Vulnerability Index. At each site, we deployed four types of natural litter bags: monocultures of 1) gingko (Gingko biloba), 2) sweetgum (Liquidambar styraciflua), 3) red oak (Quercus rubra), and 4) sycamore (Platanus occidentalis). These species range in carbon quality, with ginkgo and sweetgum having more labile constituents than red oak and sycamore (Table 1). We also deployed tea bags containing green and rooibos tea following a modified version of the Tea Bag Index (Keuskamp et al. 2013). We anticipated that increasing temperatures and heat vulnerability would increase decomposition rates for all substrates, with labile substrates decomposing faster than recalcitrant substrates. Additionally, we expected that the absolute rates of decomposition would differ between natural leaf litter and tea bags (Didion et al. 2016); however, the relative effects of substrate quality and temperature, or heat vulnerability, would be comparable. Methods Site Descriptions To evaluate the relative effects of urbanization on leaf litter decomposition and carbon mineralization, we established four field sites across the Philadelphia urban heat gradient. To validate this heat gradient, we deployed temperature loggers (HOBO Pendant MX Temperature/Light Data Logger) at each site for the duration of the study. Philadelphia is one of the largest cities in the United States, with a metro area spanning 348 km2 and a population of 1.55 million people. The city is located within a temperate climatic zone with a mean annual temperature of 12.7°C and a mean annual precipitation of 1,119 mm. Bog’s Alley (hereafter, BA) is a narrow concrete alleyway with a thin layer of accumulated soil located behind a block of rowhouses in south Philadelphia (39° 55’ 56.64” N, 75° 11’ 6.36’ W). While the site is generally devoid of larger vegetation, there were two large trees of heaven (Ailanthus altissima) located within the center of the alley, which attracted numerous invasive spotted lanternflies (Lycorma delicatula). BA represents the most extreme levels of urbanization within our study, with the highest mean temperature (23.6°C), occurrence of non-native species, and accumulation of pollutants (e.g., plastics). Additionally, BA is located within Census Tract 32, which ranks high (hottest 20% of tracts) on the Philadelphia Heat Vulnerability Index. The Summer Winter Community Garden (hereafter, G) is an urban garden located in Powelton Village, just north of Drexel University’s campus (39° 57’ 32.80” N, 75° 11’ 20.60’ W). Specifically, we placed our study within a large ~2 m x 1.5 m plot filled with soil that had previously contained vegetables (e.g., celery) and various native and non-native annuals. This plot had limited canopy cover, but a single White Mulberry (Morus alba) tree that stood approximately 3 m from the plot and provided some afternoon shade. The G site is located within Census Tract 90, which ranks very low on the Philadelphia Heat Vulnerability Index. Woodland’s Cemetery and Arboretum (39° 56’ 54.3” N 75° 12’ 09.8” W) is a historic cemetery and green space originally built in the 1700’s along the west bank of the Schuylkill River just south of the University of Pennsylvania campus. Woodlands (hereafter, W) contains an assortment of native, exotic, and invasive plant species. Specifically, the location we used for this study was covered by orchard grasses (Dactylis spp.) and Poverty oat grass (Danthonia spicata). Additionally, canopy cover was provided by nearby Norway maple (Acer platanoides) and common hackberry (Celtis occidentalis) trees. W is located within Census Tract 369, which ranks very low on the Philadelphia Heat Vulnerability Index. Cobb’s Creek Park (hereafter, CC) is an 851-acre urban park that features extensive riparian habitat running adjacent to Cobb’s Creek on the western border of the Philadelphia metropolitan area. The park contains multiple diverse habitats, including restored freshwater wetlands and forested habitats. Specifically, the location used in our study was in forested riparian habitat approximately 10 m from the bank of Cobb’s Creek (39° 57’ 07.6” N 75° 15’ 05.0” W). The site was covered with natural debris (e.g., sticks and leaves) as well as wildrye (Elymus spp.), and American pokeweed (Phytolacca americana). Canopy cover was provided by nearby chestnut oak (Quercus bicolor) and American sycamore (Platanus occidentalis) trees. CC is located within Census Tract 82, which ranks high on the Philadelphia Heat Vulnerability Index. Despite this high Heat Vulnerability Index, the interior regions of the park, where we placed the study, create a forested microclimate that likely reduced local heat stress. Natural leaf litter decomposition bags To evaluate the effects of heat vulnerability and leaf quality on litter decomposition, we constructed 80 leaf litter bags (5 cm x 5 cm, length x height) using 1 mm mesh. This mesh size allowed for the entry of mesofauna such as mites, nematodes, pot worms, and springtails, as well as juvenile earthworms and millipedes (Swift et al. 1979). All leaves for litter bags were collected in Fall 2023 on Drexel University’s campus (39° 57’ 16.1”N 75° 11’ 17.8”W) and in green spaces in northern Philadelphia (40° 5’ 37.5”N, 75° 8’ 19.4”W and 40° 5’ 37.7”N, 75° 8’ 1.8”W) and sorted by species before being left to air dry to a constant weight. We focused on four main species: 1) gingko (Gingko biloba), 2) sweetgum (Liquidambar styraciflua), 3) red oak (Quercus rubra), and 4) sycamore (Platanus occidentalis). We targeted these species because they represent species with distinct carbon quality, with gingko and sweetgum being composed of more labile, and less recalcitrant, carbon than sycamore and red oak (Table 1). Additionally, their elemental stoichiometry differs, with gingko (53.2 ± 11.4; mean ± SE) and sweetgum (50.9 ± 13.31) having similar carbon to nitrogen (N) ratios, compared to a lower ratio for sycamore (31.0  5.36) and a higher ratio for red oak (80.9 ± 10.82). These four species also represent a broader community-level shift in street trees through Philadelphia’s recent history. Specifically, sycamore and red oak are historically dominant species along Philadelphia’s streets, but the Pennsylvania Horticultural Society rarely plants these species in Philadelphia today (LaNoce et al. 2022). Meanwhile, gingko and sweetgum are species that rose in abundance across the city due to targeted planting by the Fairmount Park Commission from the 1930s through the 1950s (LaNoce et al. 2022). In each litter bag, we added 4 g whole leaves of either gingko, sweetgum, red oak, or sycamore. We deployed five litter bags of each leaf species (i.e., gingko, sweetgum, red oak, or sycamore) on the soil surface at each of our four field sites and let them decompose for six months (May – November 2024). When deploying and collecting litter bags, we placed each individual litter bag in its own plastic bag to allow us to account for any potential litter loss that occurred during transport. At the start of the study, we also ‘deployed’ five handling controls per leaf species at each site to account for impacts of transport and handling during deployment on initial leaf litter biomass and observed minimal biomass loss. These handling controls were also used to assess the initial organic matter (OM) content and C: N of all leaf species (see detailed methodology below). Upon collection of all litter bags, we rinsed the remaining leaf litter to remove any soil or debris before drying the litter in the oven at 65°C to a constant weight. We calculated the mass loss of leaf litter (initial biomass – final biomass) then divided the total mass loss by the total days that the litter bags were deployed to obtain the decay rate (mass loss day-1). All dried litter (including from handling controls) was ground using electric grinders (Cuisinart® Coffee Grinder, Model no. DCG-20), and then ashed in a muffle furnace (four hours at 550˚ C) to estimate litter OM via loss-on-ignition. Additionally, we analyzed samples for total carbon and total nitrogen content (%) using a Carlo Erba 1112 Flash Elemental Analyzer operated by the Patrick Center for Environmental Research at the Academy of Natural Sciences in Philadelphia. Tea bag decomposition bags To further evaluate the effects of carbon quality on decomposition rates across heat vulnerability gradients, we assessed the decay rate of green tea (Lipton Green Tea Sencha) and rooibos tea (Lipton Rooibos Tea) at each of our four sites using a modified version of the Tea Bag Index (TBI; Keuskamp et al. 2013) Specifically, we constructed 48 dual-compartment leaf litter bags (5 cm x 5 cm, length x height) using 1 mm mesh then placed one green tea bag and one rooibos tea bag in each litter bag (one tea bag per compartment). The tea bags themselves are made of 0.25 mm synthetic mesh. This non-standard approach was utilized to minimize the loss of replicates in our public field sites. We deployed nine dual compartment litter bags containing tea bags at all four sites in May 2025. We then retrieved three dual compartment litter bags every two months (July, September, November) until the study concluded. When collected, we placed each dual compartment litter bag in a separate plastic bag to account for any tea loss during transport. In the laboratory at Drexel University, we removed tea bags from the