Green Infrastructure: Trees can be an important tool for retention and detention of stormwater runoff. Trees provide additional benefits, including cleaner air, reduction of heat island effects, carbon sequestration, reduced noise pollution, reduced pavement maintenance needs, and cooler cars in shaded parking lots.

Contribution of tree leaves, seeds, and flowers to phosphorus in urban runoff

While it is known that tree leaves, seeds, and flowers that fall on impervious surfaces contain phosphorus (P), and some of that P leaches out and contributes to the nutrient load in urban runoff, the proportion of P in urban runoff that comes from trees varies greatly and is unclear. Hobbie (2013, personal communication) estimated from data collected in subwatersheds of the Twin Cities, Minnesota, that the amount of P in leaf litter on streets is equal to about 40 to 60 percent of the amount of P that is exported in runoff during the warm season (May through September). However, Hobbie et al. (2013) have found that leaf litter sometimes immobilizes P. This means that if it was possible to time street sweeping to occur after leaf litter had immobilized P, leaf litter could actually decrease total P load in runoff.

The relationship between leaf litter and phosphorus in stormwater is complex and dynamic. Below is a summary of some findings from research conducted on the topic.

  • Hobbie et al. (2013) compared decomposition of leaf litter of five street tree species in a parking lot gutter in St. Paul, Minnesota (Acer platanoides L. (Norway maple), Acer x freemanii (Freeman maple), Fraxinus pennsylvanica (green ash), Quercus bicolor (swamp white oak), and Tilia cordata (little leaf linden)). The researchers found:
    • litter decomposed about twice as fast in the gutter compared to previous studies in nearby natural areas;
    • litter P cycling did not seem related to litter mass decomposition, and litter P increased and decreased unpredictably over the year; and
    • litter had the potential to lose 27 to 88 percent of initial P in 24 hour lab studies via leaching.
The researchers concluded “… our results indicated significant differences between nutrient dynamics in the street and in the laboratory leaching study. Litter of some species that exhibited substantial leaching losses of N and especially P in the laboratory did not exhibit such losses in the street, likely because of dry conditions during the first 3.5 weeks of deployment... In fact, some species that lost substantial P in the laboratory study exhibited a period of increase in P mass in the street. Notably some litter types still retained a high proportion of their initial P 1 year after the litter bags were deployed in the street even though they exhibited substantial P losses initially or after an early period of immobilization (e.g. Quercus bicolor, Acer platanoides). These results suggest that there is a high capacity for leaf litter decomposing in the street to immobilize P from the environment, likely as microbes take up P to meet nutritional demands for breaking down and using the organic matter as an energy source. Leaf litter P mass, and to a lesser degree N mass, were dynamic—increasing and decreasing throughout the study period—although the causes and timing of these dynamics are unclear. The dynamics did not relate to precipitation patterns, as changes in N or P (either in total mass or concentration) between any two harvests were not correlated with either the cumulative, daily mean, or daily maximum precipitation during that period, regardless of whether the time during which precipitation fell as snow was included in the analysis. … Thus, leaching losses triggered by rain events appear to have been offset by P immobilization into litter by decomposers between rain events." … “careful selection of street tree species and timely removal of litterfall have significant potential to reduce nutrient fluxes from streets to storm drains, particularly for P.”
  • Allison et al. (1998) conducted a monitoring study of gross pollutant loading near Melbourne, Australia. The researchers
    • found that the potential nutrient contribution of stormwater leaf litter (greater than 5 millimeters) was about 2 orders of magnitude smaller than the typical nutrient loads in urban stormwater;
    • concluded “removing leaf litter from urban waterways will do little to reduce the total stormwater nutrient load”; and
    • recommended that even though leaf litter contributed little to stormwater runoff nutrient loading, “because of their large volume, leaf litter and plant matter may need to be considered when designing litter trapping devices, and where they could cause blockages or smother aquatic habitat.”
  • Kalinosky et al. (2012) conducted a 2 year street sweeping study that investigated carbon, nitrogen and phosphorus content in fine (<2 millimeter) organic, coarse (>2 millimeter) organic, and soluble fractions of street sweeper waste material along streets with varying canopy cover and street sweeping frequency in Prior Lake, MN. Results showed that coarse organics contain 40 to 97 percent of the nitrogen and up to 87 percent of the phosphorus load in sweeper waste each month.
  • Interim results from a study by Kalinosky et al. (2013) indicate the researchers are finding greater P removal from residential streets with higher canopy cover than those with lower canopy cover.
  • Dorney Jr. (1986) studied phosphorus leaching from freshly fallen, dry tree leaves and found:
    • urban tree leaves contained a relatively high amount of total P and only a small fraction leached from entire leaves in 2 hours; and
    • an average of 148 micrograms per gram (air-dried weight) of P was leachable from entire leaves in 2 hours, representing 9.3 percent of the total P available. The amount of leachable and total leaf P varied significantly among tree species but was not affected by tree diameter. The author contends that although tree leaves contain high amounts of phosphorus, it was unknown at the time of his article whether or not they actually contributed a significant amount of total P to stormwater runoff. Nevertheless the author recommends that since “it is apparent that urban street tree leaves contain fairly high levels of total P, a portion of which is readily leachable and could contribute to P in urban would be prudent to collect for disposal urban street tree leaves as rapidly as possible...”
  • Cowen and Lee (1973) tested P leaching in laboratory columns containing oak and poplar leaves collected in Madison, Wisconsin. The researchers found a potential for leaf litter to release significant soluble reactive P in a lab, but gives no clear indication of order of magnitude contribution on a watershed scale in field conditions. Other conclusions include the following:
    • most of the soluble phosphorus leached was reactive in a molybdenum blue analysis;
    • the leaves tested yielded 54 to 230 micrograms of P per gram of leaves;
    • consecutive leachings of an oak leaf sample yielded soluble P in amounts related to the effective soaking period between leachings and to the number of preceding leachings;
    • cut up leaves released almost three times as much soluble P as intact leaves;
    • leaves collected from the littoral zone of Lake Mendota leached less P than control leaves collected on the shore nearby; and
    • the moisture retained on leaves after a rainstorm contained significant soluble reactive P.

Types of street sweeping equipment available

There are three general categories of street sweepers: mechanical broom, regenerative air, and high-efficiency sweepers. Schilling (2005a) provides a description of these practices (see pages 5 through 9 of the report) as well as a discussion of general cost range and concise lists for the advantages and disadvantages of each of these types of sweeper. Appendix A of the report by Schilling also lists sweeper manufacturers, available models, and common specifications for street sweepers.

Effectiveness of street sweeping to reduce P in urban runoff

Information: For more information, see the University of Minnesota website on quantifying nutrient removal by street sweeping

Since the 1970’s, many studies have investigated the effectiveness of street sweeping to reduce stormwater nutrient load, but results of those studies vary widely. Moreover, most studies investigating effectiveness of street sweeping to reduce stormwater nutrient loads from trees do not directly measure effects of street sweeping by monitoring P concentration in runoff before and after sweeping. Based on the results of Hobbie et al. (2013), studying P in urban runoff before and after street sweeping would more accurately indicate how much P is removed from the total P in runoff through street sweeping.

Several recent literature reviews provide an in depth review of street sweeping pollutant removal efficiency (e.g. Center for Watershed Protection 2006, Law et al. 2008, Schilling 2005a). A sampling of monitoring studies and literature reviews that investigated the effectiveness of street sweeping to reduce stormwater nutrient load are summarized below. Many more street sweeping studies are summarized in the literature reviews listed in the references section of this report.

  • Kalinosky et al. (2012) conducted a 2 year street sweeping study that investigated carbon, nitrogen and phosphorus content in fine, coarse organic, and soluble fractions of street sweeper waste material along street sweeping routes with varying canopy cover and street sweeping frequency in Prior Lake, MN. They found seasonal loads as high as 0.74 pounds per curb-mile of phosphorus and 4.96 pounds per curb-mile of nitrogen can be removed for as little as $28 per pound in targeted sweeping events. They concluded that optimization of sweeping practices is expected to keep average costs at $40 to $100 per pound.
  • Kalinosky et al. (2013) summarizes preliminary results from a 2 year street sweeping study that investigated carbon, nitrogen and phosphorus content in fine, coarse organic, and soluble fractions of street sweeper waste material along street sweeping routes with varying canopy cover and street sweeping frequency in Prior Lake, MN. The researchers state that targeted sweeping appears to be a cost-effective strategy for nutrient reduction when compared to treatment ponds, where costs are generally higher. Sweeping was most cost effective in the spring and fall when targeted sweeping operations achieved costs as low as $18 per pound of P removed and least efficient during mid-summer and mid-winter when costs were often several hundred dollars per pound of phosphorus removed.
  • Law et al. (2008) published a report based on a literature review, a survey of street sweeping and storm drain cleanout practices in the Chesapeake Bay and data generated from monitoring. The authors found street sweeping to generally have very low P removal efficiency and concluded “Despite the high pick up efficiencies of newer street sweeping technologies such as regenerative air or vacuum assist street sweepers, current monitoring protocols are challenged to detect significant differences in sediment and nutrient pollutant loading reductions that may be achieved from street sweeping. Additional pollutant contributions from areas other than public streets and roadways provide additional pollutant loadings that are unaffected by street sweeping, thus reducing the effectiveness of this practice. Similar conclusions have been made by other researchers conducting street sweeping studies where there are many sources of variability in such field-based studies that make any potential impact from street sweeping undetectable (e.g., Selbig and Bannerman 2007). The authors presented several additional conclusions.
  • If streets are ‘too clean’ due to frequent street sweeping, the percent removal will be low. The National Urban Runoff Program (e.g. Bannerman et al. 1984) suggests that, on average, streets need to have 1,000 lbs/curb mile of SPaM (Street particulate matter) for sweepers to effectively reduce the SPaM loading.
  • “The predominance of the coarse sediment picked up by street sweepers and standard monitoring study designs for street sweeping have implications for measuring the effectiveness of street sweeping. In general, the particles that are most effectively removed by street sweepers are less effectively captured (or sampled) by automated samplers… As a consequence, the usefulness of standard monitoring protocols to determine the effectiveness of street sweeping by comparing pretreatment and treatment stormwater pollutant loads is questionable.”
  • “For a given set of assumptions and sweeping frequencies, it is expected that the range in pollutant removal rates from street sweeping for total solids (TS), total phosphorus (TP) and total nitrogen (TN) are: 9 to 31 percent, 3 to 8 percent and 3 to 7 percent, respectively. The lower end represents monthly street sweeping by a mechanical street sweeper, while the upper end characterizes the pollutant removal efficiencies using regenerative air/vacuum street sweeper at weekly frequencies.”
The authors proposed total phosphorus removal efficiencies for street sweeping:
  • Monthly mechanical sweeping – 3 percent
  • Monthly regenerative air/vacuum sweeping – 4 percent
  • Weekly mechanical sweeping – 5 percent
  • Weekly regenerative air/vacuum sweeping – 8 percent
  • Sorenson (2013) studied the performance of a regenerative-air street cleaner to reduce total particulate solid and total phosphorus loading to the Lower Charles River from three streets in Cambridge, MA. He concluded the Total P median reductions resulting from a single pass of a regenerative-air street cleaner on streets in multifamily and commercial land-use types were about 82 and 62 percent,respectively, and were similar in terms of grain size between both land-use types.

Recommended street sweeping practices and estimates of phosphorus removal associated with different levels of sweeping

Generally spring and fall are the most cost effective times of the year to sweep streets. Exact recommended sweeping frequency varies depending on the canopy cover and other factors.

Kalinosky et al. (2013a) developed a guidance manual and workshop series on street sweeping best practices, as well as a calculator tool “for predicting sediment and nutrient loads to street surfaces in urban areas based on overhead tree canopy; estimating the amount of material that can be removed and the cost of removal in targeted sweeping operations; and designing sweeping programs to meet nutrient reduction goals.”

Inputs into the calculator include:

  • Timing and frequency of street sweeping event
  • Canopy cover along street sweeping route
  • Street sweeping cost per curb mile

Based on those simple inputs, the tool estimates nutrient recovery and cost per pound of P removed (Kalinosky et al. 2013b).

Information: The tools developed by Kalinowsky et al. will be useful for stormwater managers, but the tools have not undergone sufficient review to be used for determining water quality credits that can be applied to permit conditions, such as Total Maximum Daily Load Wasteload Allocations


Related pages

The following pages address incorporation of trees into stormwater management under paved surfaces

This page was last modified on 4 January 2018, at 12:09.

Minnesota Pollution Control Agency | 651-296-6300, 800-657-3864 | Assistance | Web site policy