Study sites and sampling
Riga, Latvia (Fig. 1) is located within the boreo-nemoral climate zone (56.9489° N, 24.1064° E) and its green infrastructures amounts to 8% of downtown percentage area (Nikodemus et al., 2003). During winter, most precipitations (126 mm) fall in the form of snow and, given the 90 days per year with freezing temperatures, mostly cheap NaCl salts are extensively used for deicing the street pavement (Table 1). Soils in Riga’s downtown have evolved from Baltic Ice Lake sandy deposits and show a primarily sandy texture (>85% sand, 1-12% silt, <5% clay). Those supporting street lines of trees are weakly structured, compacted, highly heterogeneous with low to medium proportions of anthropogenic artefacts and they show a low biological activity with a maximum of 6.5% organic matter in the topsoil (Bouraoui et al., 2019), overall typical for anthroposoils (IUSS Working Group WRB, 2014). In addition, salt sludge splashing and salt-contaminated snow heap melting at the foot of trees cause salt pollution more than 20 times higher than away from pavement, especially by the end of winter (Cekstere et al., 2008, Cekstere & Osvalde, 2013). Given the high heterogeneity of latter contamination in the soils of Riga’s downtown, even within the same street section, peak Na/Cl concentrations in March can thus range between 132-1568/26-745 mg kg-1, decreasing then during the vegetation season and away from pavement. Air pollution forms another environmental issue in Riga’s downtown, with still increasing levels of NO2/NO/O3/CO/PM10 (Anonymous, 2014).
For studying salt pollution and its effects in Riga’s street greenery, seven street sites with salt contamination levels in foliage representative of the full range of Na and Cl concentrations measured at downtown roadside sites were selected (Cekstere & Osvalde, 2013). Two sites showed slight (about 900 mg kg-1), two others medium (about 4500 mg kg-1) and the last three severe (about 9000 mg kg-1) salt contamination (Fig. 1). Additionally, an eighth site in a National Botanical Garden (NBG) 20 km southeast of Riga was added to the sampling list (uncontaminated site). The selected street sites were planted with 5 to 20 Tilia x vulgaris trees, 9.4 + 0.2 m high (range: 8.1 – 11.2 m) and around 100-year-old (approximate age range: 90-110; NBG site: 18.7+ 0.4 m high trees, range: 18.0 – 19.1 m, older than 110 years). At each site, three trees with similar crown condition were randomly selected. By the end of vegetation season, soil and foliage material were collected with a view to characterize the 1) nutrient and salt contamination spectrum, 2) foliar allocation of salt contaminants and 3) structural responses in foliage.
For the analysis of soil chemistry, three 0.3 L soil cores were extracted on September 12 2014 from the topsoil (0-35cm) at 0.5-1.5 m distance from the stem of each selected roadside and NBG tree and thoroughly mixed prior to transfer to the laboratory. For the chemical and structural assessments in foliage, one unshaded branch per street or NBG tree, about 50 cm long, with around 60-70 leaves and located at 3-5 m high in the lower part of crown canopy, was pole-pruned on September 16, 2014. The average percentage area of necrosis per leaf was estimated visually in the field and the scores attributed to one out of 6 leaf injury classes (asymptomatic: 0%, starting injury: 1-5%, moderate injury: 6-15%, intermediate injury: 16-30%, severe injury; 31-50%, very severe injury: 51-100%). About 50 leaves were gathered for chemical analysis whereas disks 10 mm in diameter were excised from green tissues next to the leaf rim and apart main central vein in the 2nd or 3rd leaf from branch apex, with a view to microscopy assessments and given the location of salt injury (Fig. 2). These samples were immediately fixed using 2.5% EM- or LM-grade glutaraldehyde, buffered at pH 7.0 with 0.067 M Soerensen phosphate buffer. Back to the laboratory, they were fully evacuated prior to storage in renewed fixing solution at 4 °C, waiting for further processing.
With a view to the microlocalisation of salt contaminants within leaf tissues and cells, the remaining branch material was stored in a cooler and transferred on the same day to the Centre for Microscopy and Image Analysis (ZMB) of the University of Zurich, Switzerland. On the following two days, 2 mm disk samples excised at the aforementioned leaf center and rim location (Fig. 2) were fixed by means of high-pressure freezing (HPF) in liquid nitrogen (LN2) using a LEICA EM HPM 100 (Leica Microsystems), after 10 min evacuation in 1-Hexadecene 92% and insertion in 3 mm Al sample carrier (0.2 mm recess). The fixed leaf samples were stored in LN2 until further processing.