Fig 1.
Location map of the study site and residential catchment and various instruments installed in the stormwater outlet pipe, including A) laser flow meter, B) rain gauge, and C) autosampler. The location map of the catchment was made using open data source, freely available at http://geodata.myflorida.com/datasets/swfwmd::florida-counties in ArcGIS 10.3.1 version.
Fig 2.
(A) Rainfall amount and information about 22 sampled storm events (red triangles), (B) rainfall intensity, and (C) stormwater flow associated with storm events from May to September, 2016.
Fig 3.
(A) Loads of various nitrogen forms separated into four loading groups and (B) percentage of nitrogen forms in 22 storm events from May to September 2016.
Fig 4.
δ15N–NO3– and δ18O–NO3– in 12 selected storm events from May to September 2016.
Boxes indicate the range of the δ15N–NO3– and δ18O–NO3– values for NO3– sources according to Kendall et al. (2007), Heaton (1990), and Felix et al. (2015) as shown in S2 Table.
Fig 5.
δ15N–NO3– and δ18O–NO3– in four individuals storm events (A) event 1 (number of samples, n = 14), (B) event 8 (n = 27), (C) event 13 (n = 39), and (D) event 19 (n = 10) from May to September, 2016.
Fig 6.
(A) Overall mean percent contribution of five NO3– sources and (B) Mean percent contribution of NO3– sources in 12 individual storm events from May to September, 2016.
Fig 7.
Values of δ15N and δ13C for particulate organic N (PON) in stormwater runoff samples (blue crosses) and end-members (acorns, oak leaves, and St. Augustine grass).
Source proportions of three end-members were derived from the IsoError mixing model and are estimates of the proportion of each source to PON in the stormwater runoff samples.